Cost per page: the division that settles standard against XL

Two cartridges, two yields, and whatever the two prices happen to be on the day you look. This page gives you the division that decides between them, runs the same sum for a laser and an ink tank, and then adds the thing cost per page normally leaves out: the machine itself.

Divide the price by the declared page yield. Work in cents, not dollars. On the WorkForce WF-2930, Epson declares 150 pages for the 232 black cartridge and 380 pages for the 232XL black. Those yields are the durable half of the sum; the other half is whatever the two cost on the day you look, which is why this site will not print it. Divide each price by its yield and compare. The XL yields 153% more, so it is the cheaper cartridge whenever it costs less than 2.53 times the standard - and the sticker price, on its own, tells you nothing at all.

The division, stated once

Consumable cost per page = cartridge price / declared page yield. Multiply by 100 to get cents. Keep two decimal places until the end, because the differences that decide the question are often fractions of a cent. That is the whole method. Everything else on this page is about getting the two inputs right, knowing what the output excludes, and knowing how precise it really is.

Where your two numbers actually live

This is the part most readers get wrong, so do it first. Open the manufacturer's page for the exact cartridge you are considering. Epson's own 232XL page carries a price and a list of six compatible printers - Expression Home XP-4200, XP-4205 and XP-4255, and WorkForce WF-2930, WF-2950 and WF-2970 - and no page yield at all. The yield is published on the printer's page instead, under the Specifications tab. Epson's 522 ink bottle pages behave the same way.

That split is not an oversight. A declared yield belongs to a cartridge in a particular printer, not to the cartridge on its own: Lexmark heads each of its published ISO test reports with the printer models the test was run on. So the 380 pages above is 380 pages in a WF-2930. If you own an XP-4205, look up your own model before running this sum. The spec field is usually called "Page yield", "Replacement Ink Cartridges Yield" or "Approximate cartridge yield", and its footnote will name the ISO standard - which is what tells you whether the figure is comparable to the one you are holding it against. If you are not yet certain which cartridge your printer takes, settle that first: see the cartridge identification guide.

The ratio trap in high-yield cartridges

When a manufacturer says a high-yield cartridge prints about three times as much, that is a yield ratio. It is not a price ratio, and the two are unrelated. The XL is cheaper per page only while its price premium is smaller than its yield premium. The break-even needs no price at all until the last step: divide the XL yield by the standard yield, and that multiple is the most the XL may cost. For the 232 family in a WF-2930 the multiple is 380/150 = 2.53. Priced under 2.53 times the standard, the XL is the cheaper cartridge; over it, the standard is. Hold the two prices in front of you against that multiple rather than assuming, because the answer moves with every promotion and every retailer, while the multiple does not.

Black and color are two separate sums

Do not average a black yield and a color yield. The numbers describe different things, and averaging them produces a figure that describes no page you will ever print.

It also helps to know what a color yield actually is. Under ISO/IEC 24711 each color cartridge is tested to its own end of life and gets its own result. Lexmark's published test report for the Pro5500 records raw averages of 667 pages cyan, 549 magenta and 703 yellow, from nine cartridges of each color across three printers - 36 units including the black run. The test plainly does not stop when the first color empties; if it did, magenta's 549 could not sit alongside yellow's 703. The manufacturer then derives a single composite CMY figure - 594 in that report - and declares a rounded-down number, which Lexmark gives as 500. Epson states that it rounds its declared yields downward too. Epson's single "Color: 140" for the WF-2930 is a figure of that kind: one number standing in for three cartridges that empty at different rates. Because they do empty at different rates, one set will not deliver exactly the composite yield, and the shortest-lived color governs when you buy.

On the WF-2930, Epson declares 140 pages for the standard color cartridges - one composite figure covering cyan, magenta and yellow together. To get a color page rate, add the prices of all three and divide by 140. Treat the result as a close estimate rather than an exact figure: dividing three cartridge prices by a composite yield assumes the three empty together, and on Lexmark's published per-color numbers the set-equivalent yield is about 1.1% below the plain average, so the honest rate is about 1.1% above what that division gives. A color page also consumes black, so add your black rate to your color rate for what one color page really costs.

Then weight by what you actually print. Ninety mono pages and ten color pages a month is not the average of your two rates; it is (90 x mono + 10 x color) / 100. Because mono pages dominate, the weighted figure lands far below the naive average, and averaging the two rates will overstate your cost badly. But treat even the weighted figure as a floor rather than a forecast. Epson's declared yields are measured on the ISO/IEC 24712 suite, which is a mix of text and color pages, not a run of mono-only pages; and Epson's own WF-2930 footnote warns that "Yields may be lower when printing infrequently or predominantly with one ink color." A 90/10 month is exactly that case. Count your own mix before you compare anything.

Laser: the drum is part of every page

Toner is not the only consumable a laser eats. Take the Brother HL-L2370DW, whose supplies list names the TN730 and TN760 toners and the DR730 drum unit. Brother declares the TN760 at approximately 3,000 pages in accordance with ISO/IEC 19752. But the machine also needs the DR730 drum, rated at approximately 12,000 pages, and the drum's price has to be spread across every page as well. So the honest mono figure is (toner price / 3,000) + (drum price / 12,000), not the toner term alone. At those yields you buy roughly one drum for every four TN760 toners - 12,000 divided by 3,000 - and on Brother's own listed prices that drum line is a real component of the page, not a rounding error.

The standard-capacity TN730 is rated at up to 1,200 pages under the same ISO/IEC 19752 footnote, and carries exactly the same drum share on top. The same ratio test settles the two toners: 3,000/1,200 = 2.5, so the TN760 is the cheaper toner only while it costs less than two and a half times the TN730. The drum share is identical either way and drops out of that comparison entirely.

Check your own model's supplies list rather than carrying these part numbers across. Brother names twelve printers as taking the TN760 but only nine as taking the DR730; three models on the toner list - the MFC-L2690DW, MFC-L2730DW and MFC-L2717DW - do not appear on the drum's list. A neighboring model in the same range can take a different drum, and the drum is the most expensive consumable on this machine to get wrong. Note too that the HL-L2370DW ships with a starter toner rated at approximately 700 pages, not a full TN730 or TN760, so the first cartridge you buy will not be the first cartridge you used.

Drum ratings are also softer than toner ratings, and there is a reason for it. Lexmark states that there are currently no ISO/IEC standards for testing or reporting the life of imaging units, and that it rates them by its own internal method instead. That is why Brother's toner footnote names ISO/IEC 19752 while its drum footnote says only "Drum yield is approx., and may vary by type of use." Color lasers carry more consumables again: several separate drums, a transfer belt and a waste toner container, each with its own yield and price. Read your model's consumables list and amortise every item on it.

Ink tank: the yield is on the printer page, not the bottle

Epson's product page for the 522 black ink bottle lists a price and no page yield at all. The yield is published on the printer's specification page instead. Epson's EcoTank ET-2800 page states that the printer uses 522 bottles and that replacement bottles yield 4,500 pages black and 7,500 pages color. Divide the black bottle's price by 4,500 for your black rate; the cyan, magenta and yellow bottles are sold separately, so add all three and divide by 7,500 for the color rate. Epson lists all four bottles at the same price as one another, which makes the color set exactly three times the black bottle. Because these yields are an order of magnitude larger than a cartridge's, the rates come out in fractions of a cent - which is the argument for the format, and the reason the machine's own price weighs more heavily here.

Epson's own marketing claim on that page is "Save up to 90% with replacement ink bottles vs. ink cartridges." That is a manufacturer claim, reported here as such, not a finding of ours. Epson also states the box holds up to two years of ink based on printing about 125 pages a month - an assumption about you, not a property of the printer.

Three machines, same arithmetic

The machine is part of the page too

Cost per page excludes the printer, and at low volume that exclusion is the entire answer. So here is the division the rest of this page leaves out: take the purchase price, divide by the number of pages you expect to print before you replace the machine, and add it to the consumable rate. Assume 100 pages a month for three years - 3,600 pages - and the machine costs more per page than the ink does on two of these three. Use the price you would actually pay, and check whether it is promotional. On 22 September 2026 the ET-2800 carried a struck-through price and a savings badge on Epson's own site, while the WF-2930 and the HL-L2370DW carried neither. Using that discounted figure instead of Epson's standing price doubles the ET-2800's crossover advantage over the Brother - a 100% error in the number this section exists to deliver.

Raise the volume and the hardware share collapses. At 1,000 pages a month over the same three years the machine price divides by 36,000 rather than 3,600, so that line falls to a tenth of what it was and, on an ink tank, all but disappears. Which means there is a crossover, and you can compute it: two machines cost the same once you have printed (price difference) / (cost-per-page difference) pages. Below that count the cheaper machine wins whatever its ink costs. Where one of the machines is monochrome, as the Brother is, the comparison can only be made on mono pages. And compute it twice when a price is promotional, because the discount expires long before the printer does. If your lifetime total is in the low hundreds of pages, consumable cost per page is nearly noise - see the light-use guide.

What the declared yield is measured against

Why your number will not match the declared one

The manufacturers say this themselves, plainly. HP states that ISO page yields "should be used only as a starting point for comparison purposes" and not to predict the actual yield you will get. Lexmark says actual yields can vary considerably higher or lower, and lists the causes: page coverage, image type, job size, duplex and tray selection, environmental conditions, cartridge handling at end of life, calibration, idle power-on time, spot color, and copy or fax mode. Lexmark also quotes the ISO/IEC 19798 standard itself acknowledging that customers do not print continuously and that a given customer's yield "may vary significantly from the yield measured by this test method." Separately, note that 4 to 5% is Lexmark's figure for the average coverage of a real black-and-white page found by industry research - it is not a specification of any ISO test pattern.

• The first cartridge is not representative. HP states that some of its inkjets use ink from the first cartridges for printer start-up, that this one-time usage is not included in the ISO 24711 test, and that some printers ship an "introductory cartridge" which may print less than a replacement - HP's packaging indicates whether one is included. Epson describes the cartridges and bottles supplied with a printer as intended for initial setup, with a portion used for priming. Brother's HL-L2370DW ships with a starter toner rated at about 700 pages.
• Printing a few pages at a time costs more than printing continuously. HP notes that with start/stop printing, inkjets use ink to keep nozzles clear, lowering yield below the ISO figure. Lexmark makes the same point for toner: one three-page job conserves yield better than three one-page jobs.
• Not all the ink comes out. Epson states that at the Ink End point some ink remains in the cartridge to stop air entering the printhead nozzles.
• Printing mostly one color skews the result. HP states that heavy use of one color lowers that color's yield and raises the others' - which is what makes the composite color figure a poor fit for lopsided printing.
• Drum life is softer than toner life, because no ISO standard covers it. Lexmark rates imaging units by its own internal method and notes that small job sizes, short-edge feeding and envelope printing all shorten drum life.

A caveat before you compare across brands

Read the footnote, not just the number. Epson's WF-2930 cartridge yields are footnoted to ISO/IEC 24711, tested in default mode printing continuously. But Epson's 522 ink bottle yields are footnoted as based on the ISO/IEC 24712 pattern with Epson's own methodology. That is the test pattern, not the full ISO/IEC 24711 procedure. Two numbers on the same manufacturer's site therefore sit on different footings, and a cross-brand comparison built on them carries that caveat. State it; do not hide it. Where a footnote names a full standard on both sides, the comparison is much stronger.

What cost per page still excludes

• Paper. Add your own cost per sheet.
• Electricity. Lasers draw a large fuser current; inkjets draw very little.
• Maintenance ink on an inkjet - the cleaning cycles between jobs, which the ISO test never incurs because it prints continuously.
• Consumable parts beyond the obvious ones: transfer belts, waste toner boxes, maintenance boxes. Check your model's own supplies list.
• Ink or toner supplied in the box, which offsets early running costs. Epson claims up to two years' ink with the ET-2800 at about 125 pages a month; the Brother ships a part-filled starter toner instead.

A blank sheet to fill in

Check it on your own equipment

These take you to the manufacturers and standards bodies the page draws on, so you can look up your exact model.

• Standards reference
  Lexmark, Understanding Print Cartridge Page Yield (PDF) (opens in a new tab)Six pages covering which ISO/IEC standard applies to which cartridge type, their 2004 and 2006 publication dates and 2017 updates, pictures of the 19752 and 19798 test pages with their coverage figures, the full list of factors that move real-world yield, and a section stating that no ISO/IEC standard exists for rating imaging units and drums. Linked from Lexmark's own live ISO page yields page. — kdr.lexmark.com
• Standards reference
  A real ISO/IEC 24711 yield test report (Lexmark, PDF) (opens in a new tab)What a declared yield is actually built from: every one of the 36 cartridges tested listed by serial number with its individual yield, the per-color raw averages and confidence bounds, the derived CMY composite, and the lower number finally declared — plus the logged temperature and humidity. The clearest demonstration that colors are tested separately and that one published color figure stands in for three. — www.lexmark.com
• Manufacturer specification
  Epson WorkForce WF-2930 specifications (opens in a new tab)A worked demonstration of where yields actually live: open the Specifications tab for the field 'Replacement Ink Cartridges Yield ISO pages', carrying separate figures for standard black, high-capacity black and color, with the footnote naming ISO/IEC 24711 and warning that yields fall when you print predominantly with one ink color — none of which appears on the cartridge pages themselves. — epson.com
• Manufacturer specification
  Brother DR730 drum unit specification (opens in a new tab)The drum's approximate 12,000-page rating, the footnote conceding the yield is approximate and varies by type of use, and — most useful — its list of compatible printers, which is three models shorter than the TN760 toner's. Open both pages side by side and compare the two lists before ordering. — www.brother-usa.com