Restaurant Food Cost Percentage: Formula & 2026 Benchmarks

Every menu price, every supplier negotiation, and most of the anxiety in a restaurant P&L traces back to one ratio: food cost percentage. This is the full reference for that number: the formula, a plate-costing walkthrough with real figures, benchmarks by restaurant type, theoretical versus actual variance, the 2026 inflation picture, and 12 ways to pull the percentage down without touching quality. If you are looking for the broader budget rule that allocates 30% to food, 30% to labor, and 30% to overhead, that lives in our guide to the 30/30/30 rule for restaurants. This page drills into the food third.

What is a good food cost percentage for a restaurant?

A good food cost percentage for most restaurants is 28% to 35% of food sales. The National Restaurant Association's 2025 Restaurant Operations Data Abstract, surveying more than 900 operators, puts the actual median at 32.0% of sales for full-service restaurants and 32.4% for limited-service restaurants, based on 2024 figures.

The range is wide because the right number is structural, not moral. A concept built on flour and cheese can run 20%. A concept built on dry-aged beef cannot, and should not try. What matters is whether your number fits your concept, your check average, and your beverage mix, and whether it is stable week to week.

The percentage is also not a grade. A sushi bar running 38% food cost on an $85 average check banks about $52.70 of gross margin per cover. A quick-service spot running a disciplined 25% on a $12 check banks $9.00. Percentages set the guardrails, but dollars pay the rent, which is why high-food-cost concepts can outearn low-food-cost ones.

Food cost is also one half of prime cost (food plus labor), the number that best predicts restaurant survival. How food cost rolls up into prime cost and net margin by segment is covered in our restaurant profit margin benchmarks. This page stays on the food line itself.

How do you calculate food cost percentage?

Food cost percentage equals beginning inventory plus purchases minus ending inventory, divided by food sales for the same period, multiplied by 100. It measures the cost of the food you actually used, not the food you bought, which is why the two inventory counts matter as much as the invoices.

Food cost % = (Beginning inventory + Purchases - Ending inventory) ÷ Food sales × 100

A worked week for a small independent:

• Beginning inventory (Monday morning count): $8,200
• Purchases during the period: $14,700
• Ending inventory (next Monday's count): $7,600
• Food used: $8,200 + $14,700 - $7,600 = $15,300
• Food sales for the same period: $46,000
• Food cost percentage: $15,300 ÷ $46,000 = 33.3%

Three details keep the calculation honest. First, include all food and nonalcoholic beverages, and track alcohol separately as pour cost (liquor typically runs 18% to 24%); blending them hides problems in both. Second, divide by food sales, not total revenue. Using total revenue that includes alcohol flatters the number and is the most common reason an operator believes their food cost is two points better than it is. Third, count on a fixed rhythm: same day, same time, same person, ideally before the big weekly delivery. A purchases-only shortcut (purchases divided by sales) swings wildly with delivery timing and tells you nothing about waste sitting in the walk-in.

How do you calculate food cost for a menu item?

To cost a menu item, list every ingredient with its exact recipe quantity, price each at the latest invoice cost, and add them up. That total is the plate cost. Divide plate cost by menu price to get the item's food cost percentage. Here is the full math on a smash burger with fries, using illustrative July 2026 invoice prices:

Sell that plate at $11.95 and the item runs $3.25 ÷ $11.95 = 27.2% food cost.

The same math runs in reverse to set prices. Target menu price equals plate cost divided by target food cost percentage. At a 30% target, $3.25 ÷ 0.30 = $10.83 is your floor; at a 28% target it is $11.61, which rounds naturally to an $11.95 menu price. Two refinements make plate costing reliable: add a Q factor, an allowance of roughly 2% to 4% of plate cost for condiments, garnish, and normal prep waste that recipes never capture, and recost on a schedule, because invoice prices move. With USDA forecasting wholesale beef up 9.4% in 2026, that $1.15 patty drifts to about $1.26, the plate to $3.36, and the same $11.95 price to 28.1%: nearly a full point surrendered by one commodity line. Once every plate is costed, our break-even calculator turns those margins into a covers-per-day target.

What is the average food cost percentage by restaurant type?

Across the industry, food cost lands near 32% of sales (National Restaurant Association, 2024 operator medians), but the healthy target varies by nearly 20 points between concepts. Pizza runs lowest at 18% to 25%. Steakhouses run highest at 35% to 45% and remain profitable. The variation is structural, driven by what the menu is made of.

Ranges are compiled from POS and restaurant-operations vendor benchmarks published in 2025 and 2026 (Toast, Lightspeed, Rezku, Restaurant365) together with the NRA's operator medians; they describe typical independents rather than chains, which buy at different scale. Note that the NRA medians sit near the top of these bands because they include nonalcoholic beverages and blend many concept types.

The practical use of this table is calibration, not aspiration. A steakhouse chasing a pizzeria's 22% will destroy its product; a pizzeria drifting to 32% is leaking money somewhere. Benchmark against your own segment, then judge menu items by the dollars each plate contributes, not the percentage alone.

What is the difference between theoretical and actual food cost?

Theoretical food cost is what your food cost should be if every plate matched its recipe exactly: it is calculated from recipe costs and your actual sales mix. Actual food cost is what the inventory formula says really happened. The gap between them is variance, and variance is where kitchens quietly lose five figures a year.

Every point of variance on that P&L is worth $460 a month, about $5,520 a year, so a 3.3-point gap is an $18,000-a-year problem on a modest independent. Inventory platforms such as Restaurant365 and MarketMan advise treating anything beyond 1 to 2 points as an investigation trigger, not a rounding error.

What fills the gap, in rough order of frequency: over-portioning at the line, prep waste and trim that never gets a second use, spoilage, unrecorded comps and remakes, invoice price creep that has not been repriced into recipes, and, less often but not never, theft. Spoilage in particular is usually a rotation failure before it is a purchasing failure; our guide to FIFO vs LIFO for restaurants covers the shelf system that keeps the oldest stock moving first. Attack variance in that order: tighten portion tools first, then run a waste log, then audit receiving and rotation, and the gap usually closes by half within a month.

What is happening to restaurant food costs in 2026?

Food costs are still climbing in 2026, just unevenly. USDA's Food Price Outlook (updated June 25, 2026) forecasts all food prices up 3.2% for the year and food-away-from-home prices up 3.6%. Menu prices were already up 3.5% year over year as of the May 2026 CPI (BLS). The real story is which categories carry the increase.

Read that table against your own menu. Beef-forward concepts are absorbing wholesale beef inflation of 9.4% while menu prices industry-wide rise only 3.6%; the difference comes out of margin unless plates are recosted and repriced. Produce-heavy menus need to watch the farm-level vegetable line, which USDA forecasts up 27.6% with a wide interval. Breakfast and brunch operators finally get relief on eggs after the 2025 avian flu spike. The operational response is cadence, not panic: recost your top sellers quarterly, give the most volatile categories flexible menu language (rotating specials, market-price callouts), and ask your distributor which contract or fixed-price programs they offer on your ten highest-volume items.

How do you lower food cost without cutting quality?

The durable way to lower food cost is to stop paying for food that never becomes a sold plate. Leanpath's kitchen research finds that 4% to 10% of food purchased becomes pre-consumer waste, and a Champions 12.3 study of 114 restaurant sites measured an average $7 return for every $1 invested in waste reduction. Twelve tactics, in the order most kitchens should try them:

1. Cost every recipe and keep the cards current. A menu costed in January is stale by July in a 3.2% food-inflation year. Recost the top 20 sellers quarterly and everything else twice a year.

2. Track theoretical versus actual weekly. One number tells you whether the problem is the market or the kitchen. Variance beyond 2 points gets investigated the same week, not absorbed.

3. Count inventory weekly. Same day, same time, same person, before the big delivery. Monthly counts let a portioning problem run for four weeks before anyone sees it.

4. Enforce FIFO rotation with dated labels. First in, first out, on every shelf of the walk-in. Most spoilage write-offs are rotation failures, not over-ordering.

5. Run a daily waste log. Weigh and record what gets thrown away and why. In the Champions 12.3 study, kitchens that measured waste cut it by an average of 26% within the first year.

6. Portion with tools, not eyeballs. Scales for proteins, scoops and ladles for sides, laminated plate spec cards with photos at the line. A half-ounce of drift on a protein served 200 times a week is real money.

7. Yield-test what you butcher and prep. A striploin invoiced at $9.00 per pound with a 70% usable yield actually costs $12.86 per servable pound. Price from servable cost, never invoice cost.

8. Cross-utilize every SKU. An ingredient that appears in only one dish is a spoilage risk on a slow week. Fold it into a second item or a special, and route trim into stocks, sauces, and staff meal.

9. Bid out your staples quarterly. Put your highest-volume proteins and staples in front of two or three suppliers, and audit invoices line by line for price creep and silent substitutions.

10. Engineer the menu by contribution dollars. Give high-margin, high-volume items the best placement, and rework or reprice high-cost items that do not sell. Percentages guide the analysis; dollars decide it.

11. Cook with the commodity map. With wholesale beef forecast up 9.4% and farm-level vegetables up 27.6% (USDA, 2026), build specials around stable categories and give the volatile ones flexible pricing.

12. Fix the channel mix on delivery orders. A 25% to 30% marketplace commission behaves like a second food cost on every order it touches (next section). Moving regulars to a flat-fee direct channel such as DirectOrders ($249 per month, zero commission, as of July 2026) deletes that line without changing a single recipe.

Notice that none of the twelve is "buy cheaper food" or "shrink the portion." Quality cuts show up on the plate and in reviews within weeks. Waste cuts show up only on the P&L.

How do delivery commissions change your effective food cost?

A marketplace commission is not food cost on the P&L, but at the order level it behaves exactly like one. On a $40 delivery order with a 30% food cost, the food costs $12.00. A 30% marketplace commission takes another $12.00, precisely the size of the entire food budget for that order. Food plus channel now consumes 60% of the subtotal: economically, your food cost just doubled.

Those rates are not hypothetical. As of July 2026, DoorDash's published Marketplace plans charge 15% (Basic), 25% (Plus), or 30% (Premier) per delivery order, and Uber Eats' tiers run 20% (Lite), 25% (Plus), and 30% (Premium). Even an entry tier changes the math: a 15% commission on that same order pushes the food-plus-channel load to 45%, the equivalent of running a 45% food cost that no menu engineering can fix.

This is the perspective shift the rest of this guide earns you. Operators will chase half a point of portioning variance for weeks, correctly, while a channel decision silently moves the same P&L by 15 to 30 full points on every marketplace order. The honest playbook has two moves: mark up marketplace menus to claw back part of the fee (knowing it pushes price-sensitive customers around), and give your regulars a direct channel where the commission line is zero. That second move is why we built DirectOrders as a flat fee, $249 per month with zero commission (as of July 2026): at 300 direct orders a month, the platform costs about $0.83 per order against the $6.00 to $12.00 a marketplace takes from the identical ticket.

Food cost percentage is the most controllable number in the restaurant. Know it weekly, cost every plate, hold variance under 2 points, reprice with the commodity map, and audit the channel line with the same energy you give the walk-in. The kitchens that do all five do not just hit their target percentage; they keep the dollars that percentage was always standing in for.

Sources

• USDA Economic Research Service, Food Price Outlook: Summary Findings (June 25, 2026 update)
• U.S. Bureau of Labor Statistics, Consumer Price Index news release (May 2026 data)
• National Restaurant Association, "Restaurant operators kept food cost ratios in check in 2024" (2025 Restaurant Operations Data Abstract)
• Champions 12.3, "The Business Case for Reducing Food Loss and Waste: Restaurants" (2019)
• ReFED, Restaurant Food Waste Action Guide
• Leanpath, "The Path to Pre-Consumer Food Waste Prevention" (EPA-hosted presentation)
• DoorDash Marketplace pricing page
• Uber Eats merchant pricing page

Methodology notes: benchmark ranges describe typical independent operators and are compiled from the sources above plus 2025 to 2026 POS vendor guidance (Toast, Lightspeed, Rezku, Restaurant365). The plate-costing example uses illustrative July 2026 invoice prices; substitute your own invoices before setting prices. DoorDash, Uber Eats, and DirectOrders pricing figures are the vendors' published rates as of July 2026.