Blog ·
Where Amazon Arbitrage Margins Are Thinnest (And Why the Biggest Category Is on the List)
Max VershininFounder of Arbitrage Heroamazon categoriesonline arbitrageroiproduct sourcing
Short answer: the Amazon categories with the thinnest online arbitrage margins are CDs and Vinyl, Clothing, Shoes and Jewelry, and Health and Household. In each of them the median deal returns between 4% and 10% ROI, and only 21% to 23% of deals reach a 30% return. The best margins sit in Industrial and Scientific, Arts, Crafts and Sewing, Baby Products and Office Products. Clothing is on the weak list even though it produces more deals than any other category. That is the whole point. A category with many deals is not the same as a category with good deals.
Most advice about picking an Amazon category answers the wrong question. It tells you where the deals are. It does not tell you what those deals are worth.
Those are two different measurements, and they do not agree. We checked ours against our own US sourcing data, and the category that produces by far the most matches also turns out to be one of the poorest by margin.
A category can have many deals and still pay badly
There are two honest ways to judge a category, and you need both.
The first is the hit rate: what share of the matches in that category make any profit at all. The second is the margin: how good the middle deal actually is, and what share of deals clear a return worth acting on. We use 30% ROI as that bar, because below that level an unexpected fee, a customer return, or a change in Amazon's price while your stock is in transit can remove the profit entirely.
A category can score well on the first and badly on the second. That combination is the most expensive one to be in, because the scan looks productive. You get a long list of results, nearly all of them technically profitable, and almost none of them worth the shipping.
How we measured this
The percentages below come from our own US matching data as it stood in mid-2026, measured across every current retailer-to-Amazon pairing in it rather than a sample. They describe the retailers we cover, so treat them as a picture of this market rather than of all of Amazon. Profit is the Amazon sale price minus what the item costs at the retailer, minus Amazon's fulfilment, referral and closing fees. ROI is that profit against what you laid out. The median is the middle deal in a category once every deal is sorted, so a single extreme result cannot move it.
Two things this does not include: your own prep and shipping costs, and any sales tax. Both come off the top, so treat every figure here as the optimistic case.
The three thinnest categories
CDs and Vinyl
The weakest category we can measure with confidence. Only 55% of its matches are profitable at all, 45% lose money outright, and the median deal returns 4%. Just 23% of its deals reach a 30% return. Physical music is a shrinking retail market, which we think contributes to discounting that is erratic rather than deep, though the data shows the outcome and not the cause.
Clothing, Shoes and Jewelry
74% of clothing matches are profitable, which sounds healthy until you look at the size of those profits. The median clothing deal returns 9% ROI, and clothing has the lowest share of deals clearing 30% of any large category we track, at 21%. 26% lose money.
This is the category that produces more matches than any other, by a wide margin. It is not a bad category to work, and it does produce more high-return deals in absolute terms than anywhere else. But you sift through roughly five results to find one worth buying, and if you judge your sourcing by how many rows a scan returns, clothing will make your results look better than they are.
Health and Household
Sits in the bottom group on every measure at once, with no compensating strength: 75% profitable, a median deal of 10%, and 23% of deals clearing 30%. Toys and Games and Video Games follow closely on nearly identical numbers.
A high hit rate is not the same as a good category
Sports and Outdoors is the clearest example of the trap, and it is worth its own section because it does not look like a problem.
On hit rate it is one of the strongest categories we measure: 85% of its matches are profitable and only 15% lose money, among the best figures anywhere. A scan of it comes back looking excellent.
On margin it sits at the bottom. Only 21% of its deals reach 30%, exactly matching clothing on that measure, though its median is a little higher at 11%. The profits are real. They are simply small. Almost everything works a little, and hardly anything works well.
If you have ever finished a sourcing session that felt productive and then wondered why the month did not reflect it, this is usually the reason.
Where the margins actually are
Industrial and Scientific is the strongest large category in our data, by a clear margin. 48% of its deals clear a 30% return and its median deal returns 28% ROI, against 9% for clothing. Only 11% lose money, the lowest figure of any big category apart from Arts and Crafts. It is a market that gets far less attention than clothing, and it rewards that.
One warning before you move budget into it. This category carries more listing and shipping restrictions than most: dangerous goods and hazmat rules, medical device claims, business-only listings, and brand gating are all common. Check that you can actually list and ship an item through FBA before you buy it, not after.
Three others are worth knowing about:
- Arts, Crafts and Sewing: the highest hit rate of any large category at 91% profitable, with only 9% losing money and 39% clearing 30%. Consistent rather than spectacular.
- Baby Products: an unusual shape. Fewer of its deals sit in the comfortable middle, but 29% return above 100%, the highest of any large category. There are fewer deals, and the profit on each one is larger.
- Office Products: unremarkable on hit rate at 79%, comfortably above average on margin with 37% clearing 30%.
This pattern is worth recognising. The categories most sellers go to first, because they are familiar and there is plenty to look at, are the ones with the thinnest returns. The categories with the best returns are less crowded, more boring, and require you to learn a market you have no personal interest in.
One category is not thin, it is broken
Amazon Renewed deserves separate mention because it does not belong on the same scale as everything else. 95% of its matches lose money and the median deal returns -18%. No other category comes close: the next worst loses money on 45% of matches.
The cause is not a thin market, it is a mismatch. Renewed listings are refurbished units. Comparing a refurbished Amazon price against a retailer's new-item price sets two different products side by side, so the margin is not a margin at all. Renewed also carries its own seller approval requirements. Exclude the category rather than working through it.
What to do with this
- Filter on ROI, not on result count. A scan that returns a long list is not a good scan. Set a minimum ROI that survives your prep costs and a fee surprise, and judge the session by what clears it.
- Set a minimum profit in money as well as a percentage. A high ROI on a very cheap item is still a small amount, and it will not cover your time.
- Do not abandon the big categories, but change what you expect from them. Clothing and Sports and Outdoors are worth working for steady flow. Just do not judge them by how much they return per scan.
- Spend deliberate time in an unfashionable category. Industrial and Scientific and Office Products reward it, and fewer people are looking.
- Exclude Amazon Renewed outright.
- Confirm the fees before you buy with an FBA calculator. Every figure above is calculated before your prep and shipping, so the real margin is thinner than the scan says.
Keep reading
- Which Amazon Categories Have Profitable Arbitrage Deals - the companion piece: which categories can never produce a deal at all.
- How to Find Profitable Amazon Online Arbitrage Deals in 3 Minutes - the filter workflow behind the advice above.
- How to Choose Replenishable Amazon Products That Sell Consistently - turning a good category into repeat purchases.
- 10 Beginner Mistakes in Online Arbitrage - the fee and pack-quantity traps that eat a thin margin.
- Is Online Arbitrage Still Worth It? - the wider case.
Frequently asked questions
Which Amazon category has the worst arbitrage margins?
CDs and Vinyl, among categories with enough products to judge reliably. Only 55% of its matches are profitable at all, 45% lose money, and the median deal returns 4% ROI. Amazon Renewed is far worse still, with 95% of matches losing money, but it is small enough that it is better treated as a category to exclude than as one to compare.
Why is Clothing a weak category if it has the most deals?
Because volume and margin are different measurements. Clothing produces more matches than any other category and 74% of them are profitable, but the median deal returns 9% ROI and only 21% clear a 30% return. You get a great deal of activity and a thin average result, so a scan feels productive while the month does not reflect it.
What is a good ROI for online arbitrage?
Many experienced sellers use 30% as a working minimum, and that is the bar used throughout this article. The reason is practical rather than ambitious: below roughly 30%, a fee correction, a customer return, or an Amazon price move while your stock is in transit can remove the margin. Many sellers find that percentages in the low teens do not cover their time and handling once prep and shipping are counted. Your own costs decide where the line sits for you.
Which Amazon categories have the best arbitrage margins?
Industrial and Scientific leads clearly, with 48% of its deals clearing a 30% return and a median of 28% ROI against clothing's 9%. Arts, Crafts and Sewing has the highest hit rate of any large category at 91%. Baby Products has an unusual shape, with fewer mid-range deals but 29% returning above 100%. Office Products is quietly above average at 37% clearing 30%.
Should I stop sourcing in Clothing and Sports and Outdoors?
No, but change what you expect from them. Both produce a lot of matches and a steady flow of work, and clothing produces more high-return deals in absolute terms than anywhere else simply because it is so large. The mistake is measuring your sourcing by how many results a scan returns rather than how many clear your ROI bar. Work them for volume, and look elsewhere when you want margin.
Why does Amazon Renewed lose money so often?
Because the comparison itself is wrong rather than the market being difficult. Renewed listings are refurbished units, so a sourcing tool matching them against a retailer's new-item price is comparing two different products and reporting a margin that does not exist. Renewed additionally has its own seller approval requirements. The practical answer is to exclude the category from your filters entirely.