Key takeaways
- Actual food cost % = (beginning inventory + purchases − ending inventory) ÷ food sales × 100.
- Theoretical food cost is what the recipes say you should have spent on what you sold; the gap is your controllable loss.
- Food takes about 33 cents of the typical sales dollar[1]; your own target depends on concept and menu mix.
- Beef was forecast to rise 8.7% in 2026 against 1.2% for poultry[2]; recost protein-heavy dishes first.
Two numbers, not one
Most restaurants track one food cost number: what they actually spent on food as a share of food sales. That number tells you whether you have a problem. It does not tell you what the problem is. For that you need a second number, theoretical food cost, which is what you should have spent if every dish had been made exactly to recipe, with no waste, theft or mistakes. The difference between the two is where money leaks.
The stakes are large. The National Restaurant Association estimates that food takes roughly 33 cents of every sales dollar, about the same as labor[1]. A two-point gap between actual and theoretical food cost on $1 million of food sales is $20,000 a year, often more than a restaurant's entire marketing budget.
Calculate actual food cost
Actual food cost % = (beginning food inventory + food purchases − ending food inventory) ÷ food sales × 100
Count inventory at the start and end of the period
Use the same day and time each week or month, and the same units and prices.
Total food purchases
Add every food invoice received in the period, less credits. Keep beverages and packaging separate.
Calculate cost of food used
Beginning inventory plus purchases minus ending inventory.
Divide by food sales
Use net food sales from the point-of-sale system, excluding beverages, sales tax and tips.
Adjust for transfers
If the kitchen sends ingredients to the bar, such as fruit and juices, or the bar sends wine to the kitchen, record the transfers so each cost lands in the right place.
Calculate theoretical food cost
Theoretical food cost needs two things: an accurate cost for every menu item, and a count of how many of each you sold.
- Cost every recipe at current invoice prices, using yields after trimming, cooking and portioning. Include garnishes, sauces, oil and bread.
- Export the product mix for the period from the point-of-sale system: units sold for each menu item.
- Multiply units sold by plate cost for each item and add them up. That is the theoretical cost of food sold.
- Divide by food sales for the same period.
- Subtract theoretical from actual. The result, in points and dollars, is your variance.
Many point-of-sale and inventory systems calculate theoretical cost automatically once recipes are loaded. The work is in keeping recipe costs current: when invoice prices change, plate costs change with them.
Example: finding a two-point gap
Example (fictional). A casual restaurant has monthly food sales of $95,000. Its actual food cost is $31,350, or 33.0%. Its recipes and product mix give a theoretical cost of $29,450, or 31.0%. The two-point gap is worth $1,900 a month.
| Measure | Amount | Share of food sales |
|---|---|---|
| Food sales | $95,000 | 100% |
| Actual food cost | $31,350 | 33.0% |
| Theoretical food cost | $29,450 | 31.0% |
| Variance | $1,900 | 2.0 points |
In the Example, the kitchen manager checks the top ten items by cost. Two burgers are being portioned at 7 ounces instead of 6, the fry station is throwing out half-full baskets at close, and comps for slow tickets are being rung as waste instead of as comps. Fixing portions and fry batching recovers about half the gap within a month.
Where the gap comes from
| Cause | Signs | Fix |
|---|---|---|
| Over-portioning | High-cost items show the largest variance | Portion scales, scoops and plating photos at each station |
| Waste and spoilage | Full bins at close; prep thrown out | Prep to forecast, first in first out, a waste log |
| Recipe drift | Dishes look different by shift | Recipe cards at stations; spot checks |
| Receiving errors | Short or substituted deliveries accepted | Check weights and counts against the invoice at the door |
| Theft | Variance concentrated in high-value proteins or alcohol | Locked storage, regular counts, limited access |
| Unrecorded comps and staff meals | Variance that matches busy service nights | Ring every comp and staff meal in the system |
| Stale recipe costs | Theoretical cost looks too good | Update plate costs when invoices change |
| Counting errors | Big swings that reverse the next period | Count in pairs, in shelf order, at the same time |
Receiving: the cheapest control you have
Food cost problems often start at the back door. A short delivery signed for in a hurry, a substituted product at a higher price, or a case of produce that spoils two days early all show up in your food cost weeks later, when nobody remembers why. A five-minute receiving routine catches most of them.
- Check every delivery against the purchase order, not just the invoice, so you notice items you did not order.
- Weigh proteins and count cases before the driver leaves; note shortages on the invoice and get a signature.
- Check prices on the invoice against your agreed price list and flag any increase to the manager the same day.
- Check temperatures of chilled and frozen products and reject anything out of range.
- Date and put away product immediately, oldest stock to the front.
Keep beverage cost separate
Combining food and beverage into one cost percentage hides problems in both. Drinks usually run a lower cost percentage than food, so a busy bar can make a struggling kitchen look healthy, and a slow bar can make a well-run kitchen look wasteful. Calculate beverage cost the same way as food, with its own inventory, purchases and sales, and split it further into beer, wine, spirits and non-alcoholic drinks if you can. Bar variances come from pour sizes, unrecorded comps, spillage and theft, and they need their own weekly count.
Setting a target
There is no single correct food cost percentage. Pizza, pasta and coffee can run low food costs; steak and seafood restaurants run high ones and rely on higher checks. The NRA figure of roughly 33 cents per sales dollar[1] is an industry-wide reference point, not a target for your concept. Set yours from your own menu mix and profit and loss statement, then track actual against theoretical every week.
Watch the price environment as well. Menu prices for food away from home rose 3.5% in the 12 months to September 2026, while grocery prices rose 1.8%[3][4]. Raising prices faster than supermarkets is harder to justify to guests, which makes controlling the variance more valuable than pushing prices up.
Recost the dishes most exposed to price rises
Not every ingredient is moving at the same speed. USDA's October 2026 outlook forecast beef and veal prices to rise 8.7% in 2026, against 1.2% for poultry and 1.0% for pork[2]. A steak, burger or short-rib dish that was costed last year may now carry a much higher plate cost than its menu price assumes.
Source: USDA Economic Research Service, 2026[2]
Show the data
| Item | Value |
|---|---|
| Beef and veal | 8.7% |
| Poultry | 1.2% |
| Pork | 1% |
- Rank your menu by plate cost and recost the top ten every month.
- Set a trigger, such as any dish whose plate cost rises by more than a set amount, for a decision on portion, garnish, recipe or price.
- Look for substitutions that guests will not mind, such as a different cut or blend, before cutting portions.
- Use menu placement to steer guests toward dishes with stronger margins.
When the answer is a price change
Sometimes the variance is small and the plate cost is accurate, but the dish still costs too much to make at its current price. Then the answer is the menu price. Raise prices on the dishes that need it rather than across the board, round to price points that feel natural, and avoid frequent small increases that guests notice every visit. Pair an increase with a visible improvement where you can, such as a better side or a fresh garnish. Track the dish's sales for four weeks after the change; if volume holds, the price was right, and if it falls sharply, rethink the dish rather than the price again.
Weekly food cost routine
- Monday: count inventory, or at least the high-cost items, and enter the week's invoices.
- Tuesday: calculate actual and theoretical food cost and the variance by category.
- Wednesday: review the top variances with the chef and agree two fixes.
- Thursday: recost any dishes whose ingredients changed price.
- Friday: check that the fixes are in place on the line before the weekend.
Keep the routine even in good weeks. With 42% of operators reporting their restaurant was not profitable in 2025[5], the restaurants that stay profitable are usually the ones that look at food cost every week.
Frequently asked questions
What is the formula for food cost percentage?
Beginning food inventory plus food purchases minus ending food inventory, divided by food sales, times 100. Use the same counting method and period every time.
What is a good food cost percentage for a restaurant?
It depends on your concept and menu mix. Industry-wide, food takes roughly 33 cents of each sales dollar[1]. The most useful target is your own theoretical food cost, with the variance kept as small as possible.
What is the difference between actual and theoretical food cost?
Actual food cost is what you spent, based on inventory and purchases. Theoretical food cost is what you should have spent on what you sold, based on recipe costs and the product mix. The difference shows waste, over-portioning, theft and errors.
References
National Restaurant Association (2026). Restaurants remain resilient despite challenging business conditions (opens in a new tab) ↑
USDA Economic Research Service (2026). Food Price Outlook - Summary Findings (opens in a new tab) ↑
US Bureau of Labor Statistics (2026). Consumer Price Index: series CUUR0000SEFV (opens in a new tab) ↑
US Bureau of Labor Statistics (2026). Consumer Price Index: series CUUR0000SAF11 (opens in a new tab) ↑
National Restaurant Association (2025). Persistent Cost Increases and Enduring Demand Will Shape the Restaurant Industry in 2026 (opens in a new tab) ↑