Profitability · 13 min read
How to calculate food cost: formula, example and targets
Food cost is the share of a dish's menu price that goes on the ingredients, and it is the difference between a full restaurant and a profitable one. Ask an owner what the room took last night and you get a number to the cent; ask what the margin was and the answer gets vaguer. This guide covers the food cost formula, a dish costed line by line, what a healthy percentage looks like by concept, and what to do when the number comes back too high.
Published August 20, 2026By the Tablerooms team
Profitability
28-34%
where a healthy food cost lands
What is food cost in a restaurant?
Food cost is what the ingredients of a dish cost you, expressed as a percentage of what you sell it for. A dish that costs $4.28 to produce and sells for $18.90 runs at 22.6% food cost. The percentage matters because it is the only number that lets you compare a starter with a steak, and it is the fastest way to see which parts of the menu are carrying the others.
There are two versions of the number, and they answer different questions. The plate cost tells you whether a dish is priced correctly. The period cost tells you whether the kitchen is running correctly. Restaurants that track only one of them are always surprised by the other.
What is the food cost formula?
Use the plate formula when you write or reprice a menu, and the period formula when you close a month. Applying your plate costs to what you actually sold gives a theoretical cost; the period formula gives the actual. Actual is always higher, and the size of the gap is the subject of its own section below.
How to calculate food cost per dish
Cost every component at its purchase price, in the portion you actually serve, then add an allowance for the small things. A grilled sea bass main, costed the way a recipe card does it:
| Ingredient | Portion | Cost |
|---|---|---|
| Sea bass fillet at $14.50/kg | 180 g | $2.61 |
| Potatoes | 150 g | $0.23 |
| Seasonal greens | 80 g | $0.54 |
| Lemon-butter sauce, batch-costed | 60 ml | $0.48 |
| Oil, seasoning, bread service | allowance | $0.42 |
| Plate cost | $4.28 |
At a menu price of $18.90 that plate runs 22.6% food cost. Price it at $16 and it runs 26.8%. Neither is wrong, but only one of them was a decision. The whole point of costing dishes is that pricing stops being a feeling.
What is a good food cost percentage?
Most full-service restaurants aim for a blended 28% to 34%, but the right target depends on your concept. Protein-heavy menus run structurally higher than pasta-heavy ones, and that is fine when the price carries it.
| Concept | Typical food cost |
|---|---|
| Café or quick service | 25-30% |
| Casual dining | 28-33% |
| Steakhouse, seafood | 32-38% |
| Tasting menus | 30-35% |
| Bar and drinks-led | 18-24% |
What matters is the menu-mix average, the blended cost of what actually sells, rather than the worst dish on the card. A 40% dish that sells twice a week is a rounding error; a 34% dish that sells two hundred times is your business.
Food cost percentage or dollar margin?
Judge every dish on both, because the percentage can point the wrong way. A rib-eye at 38% that leaves $19 of margin on the plate beats a pasta at 22% that leaves $9. You bank dollars, not percentages, and a menu optimised purely for percentage tends to drift towards cheap dishes nobody came for.
| Dish | Price | Cost | Food cost % | Margin |
|---|---|---|---|---|
| Rib-eye | $31.00 | $11.80 | 38% | $19.20 |
| Pasta | $11.50 | $2.55 | 22% | $8.95 |
How to lower food cost
Work in this order, because the cheapest fixes are the ones the guest never notices:
- 01Re-spec before repricing. Portion size, garnish, or a different cut that tells the same story for less.
- 02Cost the yields properly. A kilo of whole fish is not a kilo of fillets, and costing the invoice weight understates every plate.
- 03Close the waste loop. Log and price trim, burns and drops, so you know what is waste and what is shrink.
- 04Renegotiate the top ten lines. Ten ingredients usually carry most of your spend, and a 5% move there beats a menu rewrite.
- 05Fix the mix. Push the dishes that carry margin, using menu position and the floor team, before touching prices.
- 06Reprice in small steps, quarterly, on the dishes that still do not clear their target.
Menu engineering: stars, workhorses, puzzles and dogs
Menu engineering plots every dish on two axes, how often it sells and how much margin it leaves, and each of the four quadrants has a move:
- Stars, popular and high-margin. Protect them: keep them where the eye lands, and never let a supplier price change erode them silently.
- Workhorses, popular with a thin margin. Re-cost first, then nudge the price. A $1 move on a dish that sells 200 times a month is $2,400 a year.
- Puzzles, profitable but ignored. Rename, replate, reposition on the card, or have the floor sell them. If nothing moves the numbers, let them go.
- Dogs, unpopular and unprofitable. Cut them without sentiment, because every dog steals prep time, fridge space and menu attention from a star.
Run the grid quarterly. With sales in the same system as the recipes, the quadrants draw themselves, with no export and no spreadsheet weekend.
Why recipe cards keep the number honest
A plate cost is only as good as its inputs, and three of them drift constantly. Yields: cost the usable portion, not the invoice weight. Batch preps: the sauce has its own recipe and its own cost per litre, and dishes inherit it. Supplier prices: when butter jumps 18%, every dish touching butter should re-cost itself the same day, which happens automatically when purchase prices feed the recipe cards. A card costed in January and trusted in August is fiction.
Theoretical vs actual food cost: finding the leak
Theoretical cost is what the recipes say you used, multiplied by what you sold. Actual is what the stocktake says left the building. The difference is your leak, and it splits in two: waste you logged, priced in a waste log, and the unexplained remainder, which is over-portioning, unrecorded prep loss, or the back door.
When should you reprice the menu?
- An ingredient moves 10% or more and stays there: re-cost every dish it touches that week, not at the annual menu reprint.
- A dish drifts three points above its target: re-spec before you re-price.
- Quarterly, regardless. Small and often beats big and rare, because guests notice a $2 jump after two years but not $0.50 a quarter.
- Use structure to steer. A well-built set menu moves guests towards margin while reading as generosity, provided you cost it on the worst-case path.
Frequently asked questions
- What is the food cost formula?
- For a dish: ingredient cost of one serving divided by menu price, times 100. For a period: opening inventory plus purchases minus closing inventory, divided by food revenue, times 100. Use the first to price a menu and the second to check the month.
- What food cost percentage should I aim for?
- Most full-service restaurants land at a blended 28% to 34%. Protein-heavy concepts run higher and compensate on price, while drinks-led venues run much lower. Set a target per dish category rather than one number for the whole card, and watch dollar margin alongside the percentage.
- How do I calculate food cost per serving?
- Cost every ingredient at its purchase price in the portion you actually serve, using the usable yield rather than the invoice weight, add batch preps at their own cost per litre or kilo, and add a fixed allowance for oil, seasoning and garnish. Divide the total by the menu price.
- How often should I run a stocktake?
- Monthly at minimum, and weekly for high-volume or high-shrink categories such as proteins and spirits. Count blind, without the theoretical number on the sheet, and price the gap every time. A count that takes a whole night usually means the storeroom needs organising more than the count needs skipping.
- Do staff meals and comps count as food cost?
- They inflate your actual cost if you leave them unrecorded, and then you chase a leak that is really lunch. Log them as their own categories, so the gap you investigate is genuine waste and shrink rather than hospitality.
- How do I cost a set menu or tasting menu?
- Cost the worst-case path, meaning the most expensive choice at every course, and check the expected mix against real sales after a few weeks. If the worst case clears your target then every combination does. If it only works on the average, one popular expensive course can sink the format.
- My supplier prices jumped. Should I raise prices or cut portions?
- Re-spec first: portion size, garnish, or a different cut, meaning changes that preserve the dish's story. Then reprice in small quarterly steps without announcements. Guests forgive quiet dimes far more easily than a headline jump, and a dish that can no longer carry its cost at a fair price is a candidate for removal rather than a worse version of itself.
- What is the difference between food cost and prime cost?
- Food cost covers ingredients only. Prime cost adds labour, and is the number most operators actually manage to, with a common target of 55% to 65% of revenue. Food cost tells you whether the menu is priced right; prime cost tells you whether the business works.