Is Costco Worth It for Two People? We Did the Math

Nearly every article on this question traces back to one figure. In a supermarket price study published in March 2026, Consumer Reports found Costco's prices ran 21.4% below Walmart's — the widest gap of any chain it measured, ahead of BJ's at 21%, Lidl at 8.5%, and Aldi at 8.3%. Whole Foods came in nearly 40% above.

It's a solid piece of research, and it's worth understanding exactly what it measured. Prices were collected in person from store shelves by the research firm Strategic Resource Group in late summer 2025, inside a 48-hour window in each metro area, covering packaged goods, produce, and meat. Loyalty-card discounts were counted; manufacturer coupons weren't. And — this is the detail that matters most — the pack sizes varied by what each store happened to carry.

That last point is not a flaw in the study. It's simply the difference between two kinds of question. A shelf-price index tells you what a pound of something costs at one store versus another. It cannot tell you whether a household bought a pound or bought six, or whether the other five made it out of the refrigerator. Nothing in the design was built to answer that, and it shouldn't be asked to.

Which leaves the interesting question wide open. If you've ever wondered how much of your grocery bill is decided by where you shop versus how much you buy at a time, our look at which grocery stores offer the best nutrition per dollar takes the first half of that. This piece takes the second.

Here's something that surprised us, buried in a government footnote.

Every month, the USDA publishes what it costs to feed someone at four spending levels — Thrifty, Low-Cost, Moderate-Cost, and Liberal — assuming every meal and snack is made at home. But those numbers are calculated per person inside a four-person household. Smaller households are told to adjust upward. The instruction from the USDA's Food and Nutrition Administration reads: one-person households add 20%, two-person households add 10%, three-person add 5%. Four people is the baseline. Five or more subtract.

Sit with that for a second. The federal government's own food budget model contains an explicit penalty for being a household of two.

Put July 2026 numbers to it. For a couple aged 51 to 70, the monthly figures come to $269.00 and $299.50 on the Low-Cost plan. Apply the 10% adjustment and the household lands at $625.35 a month. On the Moderate-Cost plan it's $777.04, and on the Liberal plan $933.90. The adjustment by itself — the pure cost of being two people rather than four — runs between $56.85 and $84.90 every month, which is somewhere from $682 to $1,019 a year.

What that penalty looks like in practice varies a lot by where you live, which we mapped in the real cost of eating healthy in every state.

Compare that to the $65 membership fee and the proportions get strange. The small-household penalty is roughly ten to fifteen times the thing everyone spends the article arguing about. Two people already pay more per head to eat, before any membership question comes up at all. And a warehouse club is, in theory, precisely the fix for that — buy in the quantities the four-person household buys in, pay the four-person household's unit price.

In theory.

The reason it stays theoretical has a price tag, and the Environmental Protection Agency put a fairly precise one on it in April 2025. Consumer-level food waste costs about $728 per person per year — roughly $14 a week, or about 11% of what an American spends on food. In weight, it's 256.2 pounds a person.

For two people, that's $1,456 a year and something over 500 pounds of food. It's more than twenty times the Gold Star membership fee. Whatever else is true about warehouse shopping, the waste line is a much bigger number than the fee line, and it's the one that never appears in anybody's break-even math.

Where the money goes is the part that makes this a health story and not only a budget one. Ranked by cost, the biggest category of wasted food is meat, at $133.48 per person a year. Then other dairy at $74.32, processed fruit at $65.86, fish and seafood at $61.33, and fresh vegetables at $59.49. The EPA is careful to note that the foods wasted in the greatest amounts aren't the ones with the greatest savings opportunity — added sugars top the list by weight — but meat and dairy lead by a wide margin on cost.

Those five categories come to just under $790 a year for a two-person household. They are also, almost exactly, the protein and produce we spend the rest of the year encouraging people over 40 to eat more of. Which means the bulk-buying question isn't really about money at all. It's about whether buying more protein and more produce results in eating more protein and more produce, or in throwing more of it away.

We wanted a calculation you could re-run yourself, so every input is either published or printed here.

The Costco side. Costco lists warehouse prices on its own website, labeled "Warehouse-Only," with a note that delivery costs more and prices vary in Alaska, Hawaii, and Puerto Rico. We captured those on September 9, 2026. Using Costco's own posted figures avoids the same-day-delivery markup and gives everything a date.

The supermarket side. The Bureau of Labor Statistics publishes average retail prices for specific foods across US cities, updated monthly. July 2026 was the latest month available. This is a national average rather than one competing store, which cuts both ways — it's more representative than a single price tag, and less precise about your particular neighborhood.

The consumption rate. This is the number nobody else states, so here's ours in full: two adults, three chicken dinners a week, four ounces of raw chicken each. That's 1.5 pounds of chicken breast a week. If your household eats more or less, substitute your own figure and the whole model re-runs.

The shelf life. Storage windows come from FoodKeeper, the open dataset published by USDA's Food Safety and Inspection Service with Cornell University and the Food Marketing Institute. It covers 661 foods and is downloadable through the federal open data catalog. One caveat we'll flag rather than bury: the current release is version 128, and its data was last revised in September 2018. It remains the version the government publishes today. Where it gives a range, we used the conservative end, so nothing here manufactures waste that isn't there.

Then two pieces of arithmetic. Days to finish is pack size divided by weekly consumption. Usable fraction is the shelf window divided by days to finish, capped at 100%. And from those falls the number that turned out to matter most — the required discount, which is simply one minus the usable fraction. If your household can realistically get through 40% of a pack before it turns, that pack has to be at least 60% cheaper per pound than the small one at the supermarket just to break even. Anything less and the warehouse version cost you more.

That last figure is the useful one, because it needs no price at all. It works on any shelf tag, in any city, on any day.

Here's the model run against every Costco item we could price from Costco's own site, alongside its storage window from FoodKeeper. Prices captured September 9, 2026; national averages are July 2026.

Read down the "discount needed" column and the pattern jumps out. Nothing in a refrigerator can clear a 93% bar. Nothing in a freezer or a pantry has any bar to clear at all. The entire question turns on which of those two places the pack ends up in.

Start with the good news, because it's real and it's substantial.

Costco's published warehouse prices on the day we looked put Kirkland Signature chicken breast at $18.99 for 6.5 pounds, thin sliced breast at $25.99 for 10 pounds, tenderloin at $19.99 for 6 pounds, and canned chicken breast at $12.99 for six 12.5-ounce cans. Work those out per pound and you get $2.92, $2.60, and $3.33 respectively.

Now the comparison. Boneless chicken breast averaged $4.153 a pound across US cities in July 2026, according to the Bureau of Labor Statistics series that tracks it. That series has been remarkably steady all year — every month of 2026 has landed between $4.14 and $4.19 — which makes it an unusually clean benchmark to hold a dated warehouse price against.

So the 6.5-pound pack runs 29.7% below the national average. The 10-pound thin-sliced pack runs 37.4% below. Even the tenderloin, the priciest of the three per pound, still comes in 19.8% under. In dollars on a whole pack: about $8.00 saved on the 6.5-pounder, $15.54 on the 10-pounder, and $4.93 on the tenderloin.

That's a genuine discount, larger than the headline 21.4% index, on a category the EPA says households waste more of by cost than anything else. Everything now depends on what happens in the next hour.

At three chicken dinners a week, a two-person household gets through 1.5 pounds of chicken breast. A 6.5-pound pack is therefore about a month of dinners — 30 days.

FoodKeeper gives boneless chicken breast halves 1 to 2 days in the refrigerator from the date of purchase. Nine months in the freezer.

Put those side by side and the picture is almost comic. Left in the refrigerator, a two-person household can safely get through somewhere between 3% and 7% of that pack before the window closes. To break even at that usable fraction, the pack would need to be 93% to 97% cheaper per pound than the supermarket's. It's 29.7% cheaper. It isn't close.

Move the same pack to the freezer and the 9-month window is nine times longer than the 30 days it takes to eat. The usable fraction goes to 100%, the required discount goes to zero, and the full $8.00 stays in your pocket.

Nobody actually leaves six and a half pounds of chicken in the fridge for a month, of course. That's the point. The behavior everyone already performs without thinking about it — breaking the pack down and freezing it — isn't a nice-to-have tip at the end of a shopping article. It is the entire mechanism by which a warehouse club works for two people. The price on the sign is only an offer. The freezer is what accepts it.

Which reframes the whole membership question rather usefully. At $8.00 saved per 6.5-pound pack, it takes about eight packs a year to cover the $65 Gold Star fee — roughly eight months of chicken dinners for a couple, from chicken alone. On the 10-pound thin-sliced packs it's closer to four. If you're already thinking in terms of cooking protein in batches, our costing of batch-cooked chicken against deli meat runs in the same direction.

Freezing is one way past the shelf-life problem. Not needing a shelf life at all is the other, and it defines a whole set of categories where the two-person question simply never arises.

Canned chicken is the cleanest example in the table. FoodKeeper gives it five years in the pantry, and once opened, 3 to 4 days refrigerated — but a 12.5-ounce can is one or two meals, so the opened window never binds. Six cans is six independent decisions rather than one large commitment. Eggs behave similarly at 3 to 5 weeks refrigerated. So does hard cheese, at six months sealed.

The structural rule underneath is worth more than any individual row: a pack that subdivides into single-meal units is safe for two people; a pack that behaves as one large unit is not. Six cans, thirty-six rolls, a case of anything — the size is irrelevant because the clock only starts on the one you open. A single 6.5-pound slab of chicken is one clock, already running.

Household paper follows the same logic with no clock whatsoever. Costco's Kirkland Signature listings put paper towels at $23.99 for 12 rolls and bath tissue at $24.99 for 36. Nothing spoils. The only real constraint is where a two-person household puts 36 rolls of toilet paper, which is a storage question rather than an economic one.

The same subdividing logic is why tub dairy behaves so differently from a slab of meat, something that came up when we costed cottage cheese against Greek yogurt and skyr.

And then the non-grocery half, which the shelf test says nothing about but which frequently carries the membership on its own: fuel, the pharmacy counter, the optical and hearing centers, tires. None of it spoils. None of it scales with household size. For plenty of two-person households, that side of the warehouse is the actual case, and the groceries are incidental.

The categories that fail are the ones you'd least like to hear about, because they're the ones the health advice keeps pointing at.

Fresh produce is the hardest case, because it has no freezer escape route in any form most people will actually eat. FoodKeeper gives bagged greens 3 to 5 days refrigerated — and only 2 days once the bag is opened. Strawberries get 2 to 3 days. Mushrooms 3 to 7. Broccoli 3 to 5. Avocados 3 to 4.

Run the arithmetic on a household eating salad a few times a week, at say ten ounces of greens. A one-pound bag is eleven days of eating against a two-day window after opening: about 18% usable, a required discount of 82%. Make it a two-pound clamshell and you're at 9% usable and a 91% required discount. There is no warehouse in the country discounting spinach by 82%.

Strawberries land in the same place. A two-pound pack at a pound a week is two weeks of eating against a two- to three-day window — 14% to 21% usable, needing a 79% to 86% discount. A four-pound pack needs 89% to 93%.

None of this means the produce is bad or the price is dishonest. It means the pack was designed for a household that empties it in three days, and two people mostly can't. If produce cost is the thing you're trying to solve, the swaps we mapped in our guide to expensive healthy foods and their cheaper equivalents are a more reliable lever than pack size.

There's a second, quieter way the shelf index oversells things, and the flour row shows it. Kirkland's all-purpose flour is the organic one, at about $1.00 a pound. The BLS national average for conventional white all-purpose flour in July 2026 was $0.543. The organic whole bean coffee runs $11.75 a pound against $9.317 for conventional ground roast.

That's not Costco being expensive — it's an organic product against a conventional average, which isn't a fair fight in either direction. But it does describe something real about how a warehouse cart fills up. An index compares like with like. A shopper standing in front of one option substitutes upward, and the bill reflects the substitution rather than the discount.

The tier question is smaller than it looks, and it resolves with one division.

The Costco membership page puts Gold Star at $65 a year and Executive at $130, both plus sales tax, each including two cards. Executive adds a 2% annual reward capped at $1,250, early warehouse hours from 9 a.m., and a $10 monthly credit on same-day delivery orders over $150.

The upgrade costs an extra $65. At 2% back, it pays for itself at $3,250 of eligible annual spending — about $271 a month. That's the whole calculation. Below that number the upgrade costs you; above it, it pays. And the $1,250 cap that gets quoted so often only binds at $62,500 of annual spending, which is not a conversation a two-person household is having.

The delivery credit is the part that can swing it. Used every month, $10 back on qualifying orders is $120 a year — on its own more than the upgrade. Used never, it's zero. It comes down to whether $150-plus delivery orders are a realistic monthly habit rather than a theoretical benefit, which is a different question from how much you spend in total.

For two people, "is Costco cheaper" turns out to be the wrong question. The prices are cheaper — meaningfully so, and by more than the headline index suggests on the categories we could price. The question is whether the pack in front of you can survive your household's eating speed, and that has a specific answer for every item on the shelf.

The commodity math that circulates everywhere takes the 21.4% index, divides $65 by it, and lands at roughly $325 of annual spending. Kiplinger runs a version of that and gets $325 for Gold Star and $591 for Executive. The arithmetic is right. The input is doing all the work — because if your real net saving is 10% after spoilage, the break-even is $650, and if it's 5%, it's $1,300.

Which is why the honest answer is conditional, and the condition is behavioral rather than financial:

• A two-person household that freezes protein on the day of purchase almost certainly comes out ahead. Eight packs of chicken a year covers the Gold Star fee by itself, before fuel, pharmacy, or paper goods.
• A two-person household that buys warehouse produce on a normal weekly rhythm is very likely losing money on that portion of the cart, at any discount a warehouse realistically offers.
• A two-person household shopping mainly for fuel, prescriptions, tires, and household paper doesn't need the food math at all. Nothing there spoils.

The useful thing to carry out of this isn't a percentage. It's the one-minus-usable-fraction test: look at the pack, ask honestly how long it would take the two of you to finish it, compare that to how long it stays good, and see whether the discount on the tag is bigger than the gap. Most of the time in the freezer aisle, it is. Most of the time in the produce section, it isn't. That's the whole line, and it runs right through the middle of the store.