Full-Service Restaurant Costs: Food, Labour and Margin
The National Restaurant Association puts median labour cost for full-service restaurants at 36.5% of sales, and profitable operators closer to 34.2%. Food cost is usually quoted at 28–35%, and prime cost, meaning food and labour added together, is generally kept at or below 65%. Your own trend matters more than any of those numbers, because menu mix alone can move food cost by six points between two kitchens run with identical discipline.
Figures verified July 2026
| Metric | Published figure | Source |
|---|---|---|
| Labour cost (median, % of sales)1 | 36.5% | National Restaurant Association2026 |
| Labour cost (profitable operators) | 34.2% | National Restaurant Association2026 |
| Food cost (% of sales)2 | 28–35% | WhippleWood CPAs2026 |
| Prime cost ceiling (% of sales)3 | 65% | The Fork CPAsAccessed July 2026 |
1 Salaries and wages including benefits, full-service respondents.
2 Widely quoted, but nobody surveyed it.
3 Commonly quoted upper bound for full-service. Treat it as a guideline.
Where the money actually goes
Table service costs more because it is more service
You're paying for a whole layer of people that a counter-service business never rosters: servers, runners, bussers, someone on the door, usually a bar. It's the structural reason full-service labour sits several points above limited-service in every edition of the National Restaurant Association's reporting. Chasing a quick-service labour percentage inside a table-service room mostly just degrades the thing customers are paying you for.
Food cost is a menu question before it is a purchasing question
Two kitchens buying equally well can post food costs six points apart because one sells steak and the other sells pasta. So re-cost your ten best sellers before you start squeezing suppliers or shaving portions; that's where a single point is worth the most money, and nine times out of ten the culprit is a price you haven't moved in two years.
Prime cost is the number that actually binds
Food and labour trade against each other constantly. Buy more prepped product and your kitchen hours fall while food cost climbs; do everything from scratch and it goes the other way. Judged separately they produce bad decisions, which is why operators watch the combined figure and worry less about the split.
Covers per labour hour
Labour cost is a fraction, and nearly everyone attacks the top of it by cutting shifts. The bottom moves too. Seat a table ten minutes sooner across a full service and you've added sales to a rota you were already paying for, which improves the percentage without taking a single person off the floor.
Wait times, throughput and your cost percentages
Table turnover is what connects these benchmarks to your actual P&L. Labour cost falls as a percentage when sales rise against the same hours, and the cheapest sales in any table-service room are sitting in the gap between a table being cleared and the next party sitting down. Guests who drift off while waiting cost you twice over: the cover you never served, plus the hours you paid for regardless.
If you run this concept, the operational side of that problem is what a digital waitlist and guest paging system built for restaurants is built to solve.
Run your own numbers
Frequently asked questions
What percentage should labour cost be in a full-service restaurant?
The National Restaurant Association's 2026 State of the Industry report puts the median at 36.5% of sales, with profitable operators nearer 34.2%. Those figures include benefits, so they read higher than a straight wages-to-sales sum.
What is a good food cost percentage for a full-service restaurant?
28–35% is the range you will see quoted almost everywhere. Nobody measured it; it's a convention. Your concept moves it anyway: steak and seafood push higher, pasta and pizza sit lower. Consistency matters more than landing on a particular figure.
What is prime cost and why does it matter more?
Prime cost is food plus labour as a percentage of sales. It matters more than either half because the two trade against each other, so improving one by buying prepped product simply moves money into the other. For full-service, 65% is the usual ceiling quoted.
Why is my labour cost higher than the benchmark?
Usually one of four things: a local wage floor above the national picture, staffing that no longer matches your real covers, sales that have slipped while the rota stayed put, or a comparison figure that excludes the benefits and payroll taxes you are counting. Check what the benchmark actually includes before you conclude anything.
How do I lower labour cost without cutting service?
Raise sales against the hours you already roster rather than cutting the hours. Faster turns, fewer people wandering off while they wait, and staffing that tracks your real demand curve all improve the ratio while leaving the floor intact.
Benchmarks for other food businesses
- Quick-Service Restaurant Costs and MarginsCounter service, faster throughput, and a labour line several points below table service.
- Coffee Shop Costs: COGS, Labour and MarginBeverage COGS, a small average ticket, and a rush compressed into about two hours.
- Bar and Pub Costs: Pour Cost, Labour and MarginPour cost leads, food follows, and the wet-to-dry split decides everything.
- Pizzeria Costs and Profit MarginsCheap ingredients, expensive delivery, and a menu built on four or five commodities.
- Food Truck Costs and Profit MarginsNo rent, but commissary fees, pitch fees, fuel and a revenue line that swings wildly.
- Bakery Costs: Ingredients, Labour and WasteCheap ingredients, expensive labour, and a product that’s worthless by Tuesday.
- Takeaway and Ghost Kitchen Costs After Delivery FeesLow rent, no front of house, and a commission line that dominates everything.
- Catering Costs and Per-Head MarginsPriced per head, costed per event, and staffed for a single afternoon.
Sources
- National Restaurant Association — 2026 State of the Restaurant Industry2026 · Industry research
- The Fork CPAs — The Ideal Labor Cost for a Restaurant or BarAccessed July 2026 · Trade press — cited convention, not measured data
- WhippleWood CPAs — Restaurant Financial Benchmarks 20262026 · Trade press — cited convention, not measured data