An antique Afshar is the rug that people fall for in a shop and then assume they cannot afford, right up to the point they look at the ticket. It is a hand knotted wool pile rug from the southeast of Persia, the country around Kerman and Sirjan, drawn in hard geometric shapes on an indigo or coral ground, and it is nearly always small. That last fact is the one that governs everything else about buying one, including the price. Of the 21 antique Afshars in our Afshar collection today, the median long side is 6.2 feet. Across all 4,843 antique rugs we hold in stock, it is 10.2 feet.
This guide is about buying and identifying them, and every number in it comes from our own live stock on 11 October 2026, measured the same way on both sides of every comparison. It is the thirty fifth guide in this series and the first one where the headline price advantage turned out to be an illusion created by size, which is a more useful thing to know than any claim about tribal authenticity.
Key takeaways
- An Afshar is a small, geometric, hand knotted wool pile rug from southeast Persia, woven in the Kerman and Sirjan region. All 21 of ours are wool pile and all 21 are recorded as Persian.
- They look cheap and are not. Median $4,500 against $9,800 catalogue wide, but size for size the medians are $3,800 against $4,300, and on a wider size cut they are identical. The discount is a format artifact.
- They do not come in room size. Zero of 21 land in the room size band, against 29.5 percent of our antiques. Forty three percent are scatter size.
- Allover and geometric, not medallion. Allover 90.5 percent against 55.0, geometric 61.9 against 38.8, medallion only 9.5 against 29.3.
- Indigo and coral, not red. Blue runs 2.26 times the catalogue rate and coral 3.56 times, while red and rust is slightly under represented.
- Genuinely old. Forty eight percent are nineteenth century, against 29 percent of our antiques, and the commonest dates are around 1890 and 1900.
- The square reputation does not survive a tape measure. One of 21 is square by measurement, and two of the three we describe as square are not.
Where the name points, and where it does not
Afshar is the name of a people before it is the name of a rug. The Afshar were a Turkic speaking group within the tribal confederations of Persia, moved south and east by Safavid rulers and settled in the country around Kerman, Sirjan and Baft. The rugs that carry the name were woven there, in a landscape of villages and seasonal camps rather than in a city workshop, and that is the useful part of the label: it tells you roughly where a rug was made and in what kind of setting.
What the label does not tell you is who sat at the loom or how they lived. Our own catalogue is a good illustration. The word tribal appears in 14 of our 21 Afshar titles, at more than four times the rate it appears across our antiques. The word nomadic appears in none of them. Tribal, in a rug listing, is a description of drawing and palette, not a statement about whether a loom was portable. By the nineteenth century, which is where most surviving Afshars sit, a great deal of the weaving in that region was settled village work. If a seller tells you a rug was woven in a tent, ask how they know.
Geographically the one thing worth being firm about is the direction. Afshar country is the southeast. We have a rug in stock whose title says south Persian, whose description says southeast Persian in one sentence and northwest Persian in another, and whose type field says northwest Persian. Three directions in one record. It is a reminder that the structured fields on a rug listing, including ours, are typed by people and inherit every slip people make. If origin matters to your purchase, make the seller commit to one answer in writing.
The price story, and why the discount is not real
On a price list an Afshar looks like the bargain of the antique world. The 21 in stock run from $2,300 to $98,000, with a median of $4,500 and a middle half between $3,775 and $8,600. Fifty seven percent are under $5,000, against 26 percent across all our antiques. Against the catalogue median of $9,800 that is a 54 percent discount, and it is the kind of number a shop is happy to let you draw your own conclusions from.
The conclusion is wrong. Rug prices are driven by area more than by almost anything else, and Afshars are small, so comparing their median to the median of a catalogue full of 9 by 12 carpets compares size and calls it quality. Matched against every other antique we hold in the same size range, roughly 56 to 84 inches on the long side and excluding runners, the Afshar median is $3,800 against $4,300 for a comparison group of 866 rugs. That is 0.88 times, not 0.46. Tighten the window to 60 to 80 inches and it is 0.94 times. Widen it to 48 to 96 inches and the two medians are identical at $4,300.
So an Afshar costs what a good small antique costs. That is still genuinely useful news, because it means the entry price into real nineteenth century Persian weaving is about $4,000 rather than about $10,000, and the reason is format rather than any judgement about the weaving. It also means you should be suspicious of anyone selling you an Afshar as an undervalued secret. The market has priced it correctly for what it is.
We have run this size adjustment on every type in this series since it caught us out on the Bakhtiari guide, where a 1.61 times headline premium shrank to about 1.1 times band for band. The Bessarabian premium was the first to survive the adjustment intact. Serab is priced at the market on both measures. Bidjar carries the largest headline premium at twice the catalogue median. Afshar is the first type where the gap vanished entirely under adjustment, and it puts a question mark against Baluch, which shows the steepest headline discount of all at 0.28 times and is also almost entirely small format. That one has never been size matched and probably needs to be. The honest version of any such claim always carries its sample size with it, which is why none of ours are quoted without one.
Format: small, often long, never a room size
The single most practical fact about Afshars is what sizes they come in. Across our 21, measured from the recorded width and length in inches rather than inferred from a title, 42.9 percent are scatter size against 20.2 percent of our antiques, 33.3 percent fall in the broad area rug range, 14.3 percent are runners, and 9.5 percent are oversized. The room size band, the 29.5 percent of our catalogue that covers the classic sitting room carpet, contains exactly none of them.
If you are furnishing a living room with one rug, an Afshar is almost certainly the wrong type to be shopping. If you want something for a bedside, a study, a hall, a landing or in front of a fireplace, it is one of the best. The commonest single shape in our stock is a rug around 4 by 6 feet: look at the 4 by 6 collection or the broader small and scatter sizes and you will see how much of this weave lands there. Nine of our 21 sit in the scatter band and another seven are only a little larger.
That clustering created a measurement problem worth being open about. Our standard size bands break at a 72 inch long side, and most Afshars measure between 5.8 and 6.5 feet, so the boundary runs straight through the middle of one homogeneous group of 4 by 6 rugs. Scored on those bands the numbers point in opposite directions, and we have not published them for exactly that reason. A ratio that flips depending on where you draw a line is telling you about the line.
Three of ours are runners and they are worth noticing because they do something the type is not known for. The shrub pattern hallway runner at $6,000 is 4 by 12 feet, the multicolour hallway runner at $6,800 is 4 by 12.2, and the coral geometric runner at $9,750 reaches 4.5 by 18 feet. With only three examples that is an inventory fact rather than a claim about Afshar weaving, and we have deliberately not calculated a runner price ratio from three rugs.
At the far end sits the exception that proves the rule. The cypress tree carpet at $98,000 measures 9.2 by 24.8 feet and is more than eight times the price of the next piece in the group. Our own description of it calls the scale rarely seen and reads the piece as a commission rather than ordinary production, and we agree. It is in the collection, it is extraordinary, and no statement in this guide rests on it. The median is $4,500 with or without it.
What the designs actually are
There is a received description of Afshar design that emphasises bold central medallions. Measured against our stock, half of that is right and half is backwards. Allover drawing, meaning a field of repeated motifs rather than one dominant centre, runs at 90.5 percent against 55.0 percent catalogue wide. Geometric drawing runs at 61.9 percent against 38.8. But medallion structure, the part of the reputation people repeat most confidently, appears on only 9.5 percent of ours against 29.3 percent of our antiques, roughly a third of the normal rate. Floral drawing is also below average at 33.3 percent against 49.3.
So the honest summary is that an Afshar is a geometric allover rug. The field fills with repeated shapes, stylised flowers reduced to angles, stars, hooked forms and small animals, and it is held together by multiple narrow borders rather than by a single commanding centre. That is a genuinely different look from the city weaves of the same region: compare any of ours with the Kerman guide, where the whole composition is usually organised around a medallion or a pictorial scene.
The design that is genuinely over represented is the boteh, the teardrop form that English speakers know as paisley. Four of our 21 carry it, against 2.0 percent across our antiques, a concentration of more than nine times. Four examples is a small number and we are stating it as a fact about our stock rather than about the weave as a whole, but it is visible: the blue tribal paisley scatter at $3,750, the large scale paisley with people and animals at $4,200 and the blue paisley at $8,200 all read instantly as boteh rugs. The boteh and paisley archive is small enough to scan in a few minutes if you want to see how the motif behaves across other types.
One motif deserves a note precisely because it is so often named and is so rarely there. Descriptions of Afshar weaving routinely single out a tulip and a bird as the signature motifs. Across our 21 neither appears in the design fields and neither appears in a title. The one rug on the Afshar term carrying a bird design is a piece that is not itself titled Afshar. If a motif is being offered to you as the thing that proves a rug is an Afshar, ask to see it on the rug.
Colour: indigo and coral
Colour is the fastest informal test, and here our data is unusually clean because the prose and the structured fields agree. Blue appears on 52.4 percent of our Afshars against 23.2 percent of our antiques, and coral, salmon and pink on 33.3 percent against 9.4 percent, a concentration of 3.56 times. Seven of the 21 carry the coral term and seven say coral in the title, the same seven.
What is not there is the deep madder red people often expect from a south Persian tribal rug. Red and rust appears on one of our 21, which is 4.8 percent against 8.3 percent catalogue wide, so it is slightly under represented rather than characteristic. Ivory grounds are also uncommon here, 14.3 percent against 28.9. The combination that actually recurs is a dark indigo field carrying motifs in coral, with ivory and gold used as accents. The blue and coral scatter at $3,800 and the rustic allover blue at $4,875 are typical of it.
Because these are nineteenth century rugs, the dyes are generally natural, and that shows in how the colour behaves rather than in how saturated it is. Expect abrash, the horizontal banding that happens when a weaver moves between dye lots, and expect the back to be noticeably brighter than the front. Neither is a fault. Our guide to natural and synthetic dyes covers how to read the difference, and it matters more on a tribal rug than on a workshop piece because the dyeing was done locally and in small quantities.
The square question, and why a tape measure beats a label
Afshars have a reputation for square and squarish proportions, and it is the single claim about them most worth testing before you buy on it. In our stock the reputation does not hold. One of our 21 is square by measurement, meaning a length to width ratio of 1.10 or better, which is 4.8 percent against 6.3 percent across our antiques. Two are squarish at a ratio of 1.25 or better, 9.5 percent against 18.9 percent. On both measures our Afshars are less square than our average antique rug, and their median ratio is 1.50 against 1.46.
More instructive is where the reputation comes from. Three of our Afshars are described as square. In each case the word appears in the title, in the shape field and in the description, three separate places agreeing with each other. Two of the three are not square. The small coral piece at $2,700 measures 2 by 2.8 feet, a ratio of 1.38. The blue and red piece at $4,500 measures 4.2 by 5 feet, a ratio of 1.18. Only the allover floral piece at $4,500, at 3.2 by 3.3 feet, is actually square.
There is a general lesson in that worth more than the specific numbers. Three fields agreeing with each other is not corroboration when all three were typed from the same first impression. The only thing in a rug record that was not typed from an impression is the measurement. Ask for width and length in inches, do the division yourself, and treat every adjective as a hypothesis. That applies to our listings as much as to anyone else’s, and it is why the square rugs collection is worth browsing by measurement rather than by name. Genuinely square antique rugs are uncommon and useful, which is exactly why the word gets applied loosely.
Age: these really are old
The strongest argument for buying an Afshar is that the age is real at a price where it usually is not. Nineteenth century dating appears on 47.6 percent of our 21 against 29.3 percent across our antiques. The two commonest circa dates on them are around 1890, on six, and around 1900, also on six, each running roughly twice the catalogue rate. Two are dated to the 1850s. The mid nineteenth century ivory floral piece at $9,000 is at the older end of what we hold.
Treat any circa date as a decade rather than a year. It is an estimate drawn from structure, dye behaviour, wear pattern and drawing conventions, and a careful dealer will tell you so. Our own records show how easily that slips into false precision: one of these rugs carries a circa date of a single specific year that appears on no other rug in our entire antique stock, while its own title says mid nineteenth century. The title is the more honest of the two. If you want to understand what separates antique from vintage from new in practice, the rug age guide sets out the actual criteria.
Bags, covers and the pieces that were never rugs
Weaving in this region produced a great deal that was not a floor rug: saddle bags, animal trappings, salt bags, covers, eating cloths. Those pieces survive in numbers because they were made in numbers, and they circulate in the rug market alongside rugs. Three of our 21 are in this category or are recorded as being, and they are the most frequently misunderstood things in the group.
The saddle bag set at $6,000 is a pair of bag faces, not a rug, and our own description suggests separating and stuffing them as cushions. The horse cover at $3,000 is titled as what it is. Confusingly, the only piece in our stock carrying the horse cover type is the allover floral square piece, which is titled as a rug. The two are labelled the opposite way round from each other, which we have flagged internally and which is why neither is cited in this guide as evidence about format.
This matters commercially. A bag face is usually smaller, often more finely woven and frequently more interesting than a plain small rug of the same date, and collectors pay for them accordingly. But it was never designed to be walked on, it may have had a back removed, and it can have a seam or a closure system that will not take floor use. If you are buying for a floor, ask whether the piece began life as one.
How an Afshar sits among the other tribal and village weaves
Afshar is one of several weaves in this catalogue that get loosely grouped as tribal, and the easiest way to tell them apart in practice is by the question each one answers. Baluch rugs are the smallest and cheapest door in, 64 percent at 4 by 5 feet or under, darker in palette and more sombre in drawing. Kurdish weaving is the runner answer, with 55 percent of ours running long. Bakhtiari is the opposite, 39 percent oversized, and it owns the garden and compartment designs. Serab is almost entirely runners and gallery carpets. Malayer is the finely drawn village pile weave of west Persia, more disciplined in its drawing than any of these. Afshar is the small, square leaning, geometric allover answer, and the one most likely to be genuinely nineteenth century at a four figure price.
Against the city weaves the contrast is sharper still. A Kerman and an Afshar can come from the same hundred miles of country and share almost nothing in the way they are drawn: one is a workshop product organised around a medallion, with a drawn cartoon behind it, and the other is a village or camp product built up from repeated motifs. The legacy Kerman overview and our page on the rugs of Persia both set that regional picture out at length. If you are starting from scratch, the antique Persian rugs guide and the five minute identification checklist are the right first two pages, and the oriental rugs overview widens it out beyond Persia. For weaves from the other side of the border, the Caucasian guide covers the geometric tradition Afshar drawing is most often compared with.
What to check before you buy
The checks that matter on an Afshar are not the ones that matter on a workshop carpet. Nobody should be quoting you a knot count. We do not record one on any of these rugs and we will not invent one, and a seller who gives you a precise figure for a tribal rug has either counted it in front of you or made it up. What to ask for instead:
- Width and length in inches, then divide them yourself. This is the only objective number in most rug listings and it settles the square question in ten seconds.
- A photograph of the back. On a pile rug it shows you the knotting, the regularity of the weave and any repair that has been worked from behind.
- Close photographs of both ends and both edges. Ends and selvedges go first on a small rug that has been used, and reweaving there is common and acceptable if it is disclosed.
- Pile depth in the traffic areas against pile depth at the edge. The difference tells you how much life is left better than any condition grade.
- Whether the piece began as a rug, a bag face, or a cover.
- What the circa date is based on. A decade with a reason behind it is worth more than a year with none.
Two things we deliberately do not claim about these rugs, and that you should treat carefully wherever you see them. The first is fineness. Our material field reads wool on all 21, which describes the yarn in the visible face and says nothing about the foundation, the spinning or the knot density. If the construction matters to you, the foundation guide explains what warps and wefts do and the knots and KPSI guide explains why the number is so often misused. The second is the loom. We make no claim that any of these rugs was woven in a tent, because nothing in our records tells us that.
Living with one
A small antique pile rug is one of the easiest things to own, which is part of the appeal. Put a proper pad under it, because a 4 by 6 rug on a hard floor will creep and a pad also takes a surprising amount of the abrasion. Vacuum without a beater bar, in the direction of the pile. Keep it out of a window that gets hard afternoon sun, since on a rug of this age the dyes will go before the wool does. Rotate it end for end once a year so wear spreads evenly, which is easy on a small rug and the main practical advantage small rugs have.
Where they work best is anywhere a 4 by 6 or a 5 by 7 is the right size and character matters more than coverage: beside a bed, under a desk, in front of a fireplace, in an entry, on a landing, layered over a larger plain rug or over sisal. The runners work in a hall. What none of them will do is furnish a living room on their own, and no amount of enthusiasm about the weaving changes that.
What we have in stock, and at what price
The 21 antique Afshars we hold on 11 October 2026 give a fair picture of the range. At the entry end, the blue ground tribal piece at $2,300 and the bold graphic tribal design at $2,650 are both real nineteenth and early twentieth century weaving under $3,000. In the middle, where most of the type lives, the coral four pointed star piece at $3,800, the coral scatter at $4,400 and the paisley with people and animals at $4,200 are typical in size, drawing and palette. Higher up, the large coral piece at $9,800 and the ivory large scale tribal rug at $11,400 show what the type costs once it gets past 6 feet.
Expect to pay around $4,000 for a good 4 by 6 with honest wear, around $6,000 to $10,000 for a runner or anything appreciably larger, and more for condition, pre 1900 dating or an unusual design. If a dealer is asking a city weave price for an Afshar, the size adjustment above is the argument to make: these are correctly priced as small antiques, not as bargains, and not as fine carpets either.
Browse what is in stock now in the Afshar collection, or widen out to the tribal design archive, the Persian origin archive and the whole antique rug collection. If you are comparing types, the Baluch, Kerman and Malayer collections are the three nearest neighbours in our stock, and the cypress tree archive is where the one monumental piece in this group sits alongside the rest of our tree of life weaving.
Frequently asked questions
What is an Afshar rug?
An Afshar is a hand knotted wool pile rug from southeast Persia, woven in the country around Kerman and Sirjan by weavers of the Afshar, a Turkic speaking group resettled in that region centuries ago. The pieces that reach the market are almost all small: of the 21 antique Afshars we hold in stock today the median long side is 6.2 feet, against 10.2 feet across all 4,843 antiques we carry, and not one of them is a room size carpet. They are strongly geometric and almost always drawn as an allover field rather than around a central medallion, and indigo blue with coral is the signature palette.
Why are antique Afshar rugs so much cheaper than other Persian rugs?
Mostly they are not, once you compare like with like. Our Afshars have a median of $4,500 against $9,800 across every antique we stock, which looks like a 54 percent discount, but that gap is almost entirely about size. Matched against every other antique in the same size range, roughly 56 to 84 inches on the long side, the Afshar median is $3,800 against $4,300 for the comparison group of 866 rugs, and on a wider cut of 48 to 96 inches the two medians are identical at $4,300. In other words an Afshar costs what any good small antique costs. It reads cheap on a price list because almost none of them are big, and big is what drives rug prices.
Are Afshar rugs square?
Far less often than the reputation suggests, and it is worth measuring rather than taking a label. Of our 21, exactly one is square by measurement at a length to width ratio of 1.10 or better, and two are squarish at 1.25 or better. Across all our antiques those figures are 6.3 percent and 18.9 percent, so our Afshars are actually less square than our average antique, with a median ratio of 1.50. Three of ours are described as square in the listing and two of those three are not: one measures 2 by 2.8 feet and another 4.2 by 5 feet. The reputation is real in the sense that squarish Afshars exist and are prized, but ask for the measurements in inches before you believe the word.
What does an Afshar rug typically cost?
Across the 21 in our stock today the range runs from $2,300 for a small blue tribal piece to $98,000 for one monumental 9 by 25 foot carpet, with a median of $4,500 and a middle half between $3,775 and $8,600. Fifty seven percent of them are under $5,000, against 26 percent of all our antiques, so this is one of the few antique types where a real nineteenth century example is a four figure purchase. Runners and larger pieces climb quickly: our three Afshar runners sit between $6,000 and $9,750, and the two pieces over 12 feet are the exception that pulls the top of the range.
How do I tell an Afshar from a Baluch or a Kurdish rug?
Start with format, because in our stock the three types answer different questions. Afshars are small and almost square to moderately long, 43 percent of ours are scatter size and none are room size. Baluch are smaller still, 64 percent at 4 by 5 feet or under, and they run darker, with deep blues and browns and a great deal of camel coloured ground. Kurdish pieces are the runner answer, 55 percent of ours are runners or gallery length. Then look at the drawing: Afshar fields are allover and geometric with a strong use of the boteh, and the palette leans indigo with coral, which is uncommon in both of the others. None of these tests is conclusive on its own, which is why attribution in tribal weaving is argued rather than settled.
Are antique Afshar rugs really nineteenth century?
Many of them genuinely are. Forty eight percent of our 21 carry a nineteenth century date against 29 percent across all our antiques, and the two commonest dates on them are around 1890 and around 1900, each more than twice as frequent as in the catalogue at large. Two are dated to the 1850s. That combination of real age at a four figure price is unusual and it is the strongest argument for buying one. Bear in mind that a circa date on any antique rug is an estimate from structure, dyes and drawing rather than a documented year, so treat it as a decade and not a birthday.
Are Afshar rugs durable enough for daily use?
Yes, with the normal caution owed to any rug of this age. All 21 of ours are hand knotted wool pile, which is the hardest wearing construction a rug can have, and small tribal pieces were woven to be used. What limits them is condition rather than construction: a rug that has already been walked on for 120 years has less pile left than it started with, and low pile wears through faster at a doorway or a pivot point. Use a good pad, keep them out of direct sun, and ask for close photographs of the ends, the edges and any thin areas before you buy. We make no claim about knot count on these, because we do not record it and neither should anyone selling to you without having counted.