Search for the quantity at which screen printing becomes cheaper than DTF and you will be told 20 shirts, 36 shirts, 100 shirts, 500 shirts and 838 shirts. That is a fortyfold spread on the single number a print shop needs in order to decide what equipment to buy.
The interesting part is that this does not require anybody to be lying. The break-even quantity is calculated by dividing a setup cost by the gap between two per-shirt costs, and on published figures that gap is about four cents. A number produced by dividing by four cents is not a robust number. Move one input a little and it moves a lot, which is exactly what has happened across the industry’s published guides.
Six sources, six answers
Here is the spread, with what each publisher sells alongside it, because on this subject there is no disinterested source at all.
Jinlong Prints, a DTF supplier, says that “for runs below 20 pieces, DTF is generally more affordable than screen printing.” Arklavo, a print shop, says “DTG is cheaper under about 36 units for most designs” with screen printing winning above that. Transfer Kingdom, which sells DTF transfers, says “screen printing only becomes cheaper at high volumes, 100 plus shirts.” Gelato, a production network, agrees at “above 100 pieces on a one or two color design.” ColDesi, which manufactures and sells DTF printers, puts it at “closer to 500” for a one-colour design and “around 800” for four colours, with its calculator returning 838.
The two highest numbers come from the company selling DTF equipment, and high numbers are the ones that extend the range in which DTF is the cheaper choice. That is worth noticing and it is not by itself an accusation. What follows explains how figures that far apart can all be defensible.
Taking one model apart
ColDesi is the right one to examine, and not because it is the worst. It is the best documented. Its case study states the parameters openly: an eight by eight inch front print, 250 units, costs “updated November 2025 based on average U.S. shop data”, and production speeds taken from its own machine, which it names. Disclosing that the model is parameterised on the equipment you sell is better practice than most of this search manages.
Seven steps, and by the end the spread explains itself.
1. What ColDesi actually publishes
For that 250-shirt job, the calculator returns screen printing at $0.78 per shirt and DTF at $0.67 per shirt, making DTF cheaper by $26.67 in total, or $0.11 each. It then reports a cost break-even quantity of about 838 shirts, above which screen printing becomes the lower-cost option.
Its FAQ adds the sensitivity to colour count directly: “if you have a 4 color design and you are printing with a high-speed printer, the break-even point cost comes around 800 shirts. Alternatively, if you had just a 1 color design, the break even point would be closer to 500 shirts.” The same page also notes that screen printing “can be challenging to achieve profitability in jobs under 48 or even up to 144 pieces”, which is the same economics viewed from the other end.
Three published numbers, from one consistent model. That is enough to work backwards.
2. Reconstructing the model
Screen printing has a fixed setup cost, screens burned and press set, spread across however many shirts you run, plus a marginal cost per shirt for ink and labour. So cost per shirt is setup divided by quantity, plus that marginal cost. DTF has essentially no setup, so its cost per shirt is flat regardless of run length. That is the whole reason a crossover exists.
ColDesi gives two points on those curves: at 250 shirts screen printing is $0.78, and at 838 shirts it equals DTF’s flat $0.67. Two equations, two unknowns. Solving them gives an implied setup of about $39 and an implied marginal screen cost of about $0.62 per shirt.
Both reproduce the published figures. At 250 shirts, $39.20 divided by 250 is $0.157, plus $0.623, which is $0.78. At 838 shirts, $39.20 divided by 838 is $0.047, plus $0.623, which is $0.67. To be clear about what this is: these are figures derived from ColDesi’s published outputs, not inputs ColDesi stated. But they reproduce its numbers exactly, so they describe its model.
3. The formula, and why it is fragile
With the model reconstructed, the break-even has a compact form. The two costs are equal when setup divided by quantity, plus the screen marginal cost, equals the DTF cost per shirt. Rearranged:
Break-even quantity equals setup divided by the gap between the DTF cost per shirt and the screen marginal cost per shirt.
On ColDesi’s implied numbers that denominator is $0.67 minus $0.623, which is four and a half cents. The entire published figure of 838 shirts is $39.20 divided by less than a nickel.
This is the structural problem. When you divide by the difference between two similar numbers, small changes in either number produce enormous changes in the result. Get the DTF cost wrong by ten cents, which is well within the range of ink coverage variation on a single design, and the answer does not shift by fifteen percent. It can halve or double. The break-even quantity is a genuinely unstable quantity, and no page publishing one mentions that.
4. The input nobody agrees on
So how well established is the DTF cost per shirt? Not very.
ColDesi’s calculator uses $0.67 for an eight by eight print. ColDesi’s own FAQ, on a different page, says the cost of DTF ink and film per shirt “varies based on your ink usage, film size, and print coverage, but a typical range is 0.50 to 2.50 euros per shirt for ink and film.” CastleInk, which sells the ink, says that “most small shops land somewhere around $1 to $3 in consumables per standard 11 by 17 transfer sheet, depending on ink coverage and where you buy supplies.” Perfec Press builds it from components: ink at $80 to $120 per litre, transfer film at $1.50 to $3.00 per square foot, and labour around $25 an hour with two to five minutes per transfer.
The calculator’s figure sits at or below the bottom of its own company’s published range, and well below the ink seller’s. In fairness, an eight by eight print is considerably smaller than an eleven by seventeen sheet, so a lower number per shirt is entirely defensible. The issue is not that the figure is wrong. The issue is what section three established: the answer is extraordinarily sensitive to it.
5. Moving one number
Hold ColDesi’s implied screen printing figures exactly as they are, its $39 setup and $0.62 marginal cost, and change only the DTF consumable cost from the bottom of the published range to the middle of it.
At $0.67 per shirt, the break-even is $39.20 divided by $0.047, which is 838 shirts. At $1.50 per shirt, the break-even is $39.20 divided by $0.877, which is about 45 shirts.
Check it at 45: screen printing costs $39.20 divided by 45, which is $0.871, plus $0.623, giving $1.49 against DTF’s $1.50. They meet, as they should.
So one input, moved within the range that the industry itself publishes, takes the break-even from 838 shirts to about 45. That is the whole story of the fortyfold spread across this search. Six publishers with slightly different assumptions about ink coverage produce six answers that look like disagreements about technology and are actually disagreements about a consumable cost.
6. The other input, pulling the other way
Setup cost disagrees too, and in the opposite direction, so the range widens rather than narrowing.
Arklavo puts screen printing setup at $25 to $75 per screen. A four-colour design therefore costs between $100 and $300 to set up, against the roughly $39 implied by ColDesi’s four-colour model. Since setup is the numerator, higher setup pushes the break-even up, extending the range in which DTF wins, which is the reverse of what the consumable correction did.
Gelato describes the same economics from the shop floor. At twelve shirts, spreading setup across the run makes it “nearly $9 per piece”. At 250 to 500 pieces, a one-colour design “can drop under $3 per shirt including the blank, a number neither digital method approaches”. Both statements are about the same fixed cost behaving differently at different run lengths, which is the only thing on this subject everyone actually agrees about.
7. What actually moves the answer most
Colour count moves the break-even further than any other variable, and it is not a close contest with the others.
ColDesi’s own figures make the point: around 800 shirts at four colours, closer to 500 at one. Each additional colour is another screen to burn, another setup, another pass on the press, and more cleanup. Screen printing’s cost scales with colours. Digital methods do not care how many colours are in the file, which is why the comparison inverts completely for complex artwork. Arklavo state the corollary plainly: full-colour photographic designs “stay cheaper via DTG at any volume”, because there is no quantity at which separating a photograph into screens becomes economic.
On DTG’s own position, Gelato explain why it loses to DTF on cost for structural rather than marginal reasons: pretreatment “adds a step, a consumable, and a cure cycle to every dark garment”, and per-unit ink “runs higher than DTF, which is why DTF is considerably cheaper than DTG on comparable work.” DTG’s case is quality and hand feel on cotton, not price.
Jinlong Prints publish a cost-per-shirt comparison across one piece, 100 pieces and 1,000 pieces, which is the right shape for this decision even where the specific figures reflect their own supply prices. The pattern it shows is the one every source agrees on regardless of where they put the crossover: at a single unit, screen printing is absurd and digital wins outright; in the low hundreds the two converge and the answer depends entirely on the assumptions this article has been pulling apart; and at a thousand units on a simple design, screen printing wins by a distance that no consumable assumption can close.
That shape matters more than the crossover point, because it tells you which decisions are actually close. If you are quoting a job of twelve shirts or a job of two thousand, no calculator is needed and no published break-even is relevant. The number only matters in the band where the curves nearly touch, and that is precisely the band where a four-cent modelling difference decides the answer.
Running it on your own numbers
The formula needs three inputs and only one of them is easy to find.
Setup cost you can quote. Ring two shops or count your own screen charges, and use $25 to $75 per screen multiplied by your colour count as a starting range. Your screen marginal cost is ink plus labour per shirt at your press speed, which you can measure in an afternoon. The DTF consumable cost is the one that matters most and the one you must measure rather than look up, because it depends on your coverage, your film size and what you pay for supplies. Print a full sheet, count how many transfers you got, and divide your actual ink, film and powder spend by that number. That single measurement is worth more than any published figure in this article.
The other question the arithmetic informs is whether to own equipment at all. If your work is short runs with many colours, the crossover sits far out and outsourcing per-print means paying somebody else’s margin on every shirt indefinitely, which is the case for bringing DTF printing equipment in house. If your work is long runs of one and two colour designs, the crossover arrives early and screen printing, in house or bought in, will keep winning. ColDesi’s FAQ offers its own payback framing for the equipment decision, suggesting many small businesses break even on a printer at around 40 to 80 shirts a month depending on profit per shirt, which is a different break-even from the one this article is about and worth not confusing with it.
What this page does not cover
Quality, hand feel and wash durability are absent deliberately. Every source on this search covers them, they are substantially subjective, and we have pressed nothing and washed nothing, so any verdict here would be borrowed opinion dressed up as testing. Sublimation and vinyl are out because they carry different substrate constraints, and adding methods dilutes an argument that is already about three.
No printer model is recommended, and that applies to every manufacturer including any linked above. The whole point of the preceding two thousand words is that the answer depends on inputs specific to your shop, so a model recommendation would contradict the article.
One thing worth checking before any of this matters: fabric. Several sources note that DTG bonds poorly to polyester-heavy garments and that method choice can be settled by substrate before cost enters the conversation at all. Establish what you are printing on first, then run the arithmetic on whatever methods remain.
