Does Offer vs. Serve Reduce Plate Waste? What the Evidence Says

Sometimes. It depends on which grades you serve and on what you measure. USDA's position is that it does: its June 2025 memo SP 14-2025 says USDA research found Offer vs. Serve (OVS) "associated with significantly lower levels of plate waste for overall calories, fruits, and vegetables." The largest recent US dataset points the other way for older students. In plate-waste audits from 134 schools, schools using OVS reported 38% more waste than schools serving full trays. Middle and high schools drove that gap; elementary schools had nearly identical waste either way.

An earlier version of this article said research "consistently shows" that OVS reduces plate waste. That overstated it. The evidence is mixed, and the studies that disagree are measuring different things. Here is what OVS requires, what each study found, why they conflict, and how to find out what OVS does in your own cafeterias.

At lunch, OVS lets a student decline up to two of the five meal components and still get a reimbursable meal. Under 7 CFR 210.10(e), schools must offer all five components every day: fruits, vegetables, grains, meats/meat alternates and fluid milk. Students must be allowed to decline two, but must take at least half a cup of either the fruit or the vegetable component. In practice, that means at least three components, one of them produce.

Senior high schools must use OVS at lunch; the regulation leaves "senior high" to the state agency to define. Below senior high, OVS is at the school food authority's discretion. USDA's current guidance, SP 14-2025, calls it "optional but strongly encouraged for elementary and middle schools."

At breakfast, SP 14-2025 describes OVS as "optional but strongly encouraged for elementary, middle, and high schools." Under 7 CFR 220.8(e), a school that uses it must offer at least four food items. Students may decline one of the four, provided they take at least half a cup of the fruit component. A school that offers only three items at breakfast cannot use OVS.

The stated purpose is waste. In SP 14-2025's words, "The goal of OVS is to reduce food waste and improve menu planning in the school meal programs by allowing students to decline food they do not intend to eat." The memo goes further than the regulation and "encourages schools to use the OVS flexibility at both lunch and breakfast for ALL grade levels."

Four sources come up most often, and they do not agree. Each answers a slightly different question.

USDA's School Nutrition and Meal Cost Study is the basis for USDA's claim. A 2024 PLOS ONE paper describes its finding this way: USDA "found lower rates of plate waste for all calories and produce at 'offer' elementary schools (but not middle schools)." The USDA data were collected in 2014–15.

A 2011 study in the Journal of Child Nutrition & Management compared two elementary schools, one serve-only and one using OVS. Researchers weighed the fruit and vegetable waste (apples and bananas were estimated visually) of fourth and fifth graders over five days in October 2009. At the serve-only school, an average of 51.99% of each fruit and vegetable serving was thrown away, against 32.15% at the OVS school. Consumption per student did not differ significantly, though: 1.57 oz at the serve-only school and 1.45 oz at the OVS school. The authors note that a sample of two schools "limits the applicability of the study."

A 2019 University of Illinois Extension project at Beverly Manor Junior High, a school of about 400 students in grades 4 through 8, ran a plate-waste audit on a day it served pizza bread, pineapple and green beans. Waste was 27% of the food served. The school then introduced OVS, alongside two staff trainings, student training and extra signage. A second audit on the same menu found waste of 12.7%.

The largest recent dataset is Adjapong and colleagues (PLOS ONE, 2024). They analyzed audits that staff and volunteers from 134 schools in 24 states uploaded to World Wildlife Fund's Food Waste Warriors Dashboard between 2018 and 2023 (not all were enrolled in the program), with plate waste weighed and reported per student. Schools using OVS reported 38% more waste than serve schools. The result was "driven by middle and high schools," and elementary schools had "nearly identical waste levels regardless of offer versus serve." The authors describe the sample as "a self-selected convenience sample" and write that "we caution against drawing causal inferences."

The studies disagree mainly because they measure different things, in different grades, with designs that cannot isolate OVS. Read side by side, they conflict less than their headlines suggest.

What is measured. The 2011 study measured the percentage of each serving that was wasted. Adjapong and colleagues measured absolute plate waste per student. Those are different questions. The PLOS ONE authors suggest that students in offer schools may "take more total food" and, even if they waste "a similar or smaller fraction," produce larger absolute waste. The 2011 study shows the reverse trade-off: a lower share of each serving was wasted under OVS, but students ate about the same amount of fruit and vegetables either way. Less waste per serving is not the same as more produce eaten.

Grade level. The pro-OVS evidence is mostly elementary. USDA's finding applied to elementary schools but not middle schools, the 2011 study covered fourth and fifth graders, and the Illinois school spanned grades 4 through 8. In the PLOS ONE data, elementary schools showed no difference, and the extra waste came from middle and high schools. Because high schools must use OVS at lunch, few high schools can serve as a comparison group. In that dataset, eleven high schools reported serve mode, which the authors say may reflect a granted exception, mixed-grade schools, or "a miscoding by school volunteers."

Design. None of these studies is a controlled trial. The 2011 study compares two schools. The Illinois result is one school, one menu, one day before and one day after, with training and signage introduced at the same time as OVS, so the drop belongs to the whole package. The PLOS ONE data are observational, include no measurement before and after a school switched, and lack OVS status for nearly 12% of the sample. Practice may also have changed between USDA's 2014–15 data and the 2018–2023 audits.

None of them can tell you what OVS will do in your cafeterias, with your menu and your serving line.

At high school lunch, whether to use OVS is settled; how you run it is not. For lower grades, and at breakfast, adopting it at all is a real decision.

SP 14-2025 lists practical steps that help students and staff understand what makes a reimbursable meal: posters and signs in service and dining areas, servers and cashiers who have reviewed the menu before each service, fruit served near the point of sale so students can easily add it, and menus shared with teachers and posted on the school's website and social media. The Illinois school paired its change with staff and student training and extra signage, which fits that list.

The menu matters as much as the service model. OVS lets students decline items they will not eat, so a component that is declined or wasted day after day is a menu signal, not just a line problem. Menu changes that reduce plate waste covers portion size, rotation and recipe levers. Recipes are also about to change for reasons outside your control: USDA's weekly added-sugars limit for school meals and its lower sodium limits are currently scheduled to take effect July 1, 2027 (as of November 2026). Our guide to the 2027 USDA meal standards and plate waste covers what changes and what to watch on the tray.

Measure your own trays before and after the change, on the same menu, and record both the share of food wasted and the amount wasted per student. Mixed evidence is a reason to measure, not a reason to guess.

Fix the question first. Decide whether success means less food in the bin (pounds per student), a smaller share of each component wasted, or more fruit and vegetables eaten. The 2011 study shows these can move in different directions.

Baseline on matching menus. Measure before the change and again after it, on the same menu where you can, as the Illinois school did. Use more than one day per period, because a single day swings with the menu, the weather and the calendar.

Break it out by component and by grade. The research splits by grade, and so should your data. A district-wide average can hide an elementary improvement and a middle school increase.

Change one thing at a time, or log everything. If you add OVS, new signage and a new line order in the same week, you will not know which one worked.

Track what students take, not just what they leave. If students decline more produce, waste can fall while the amount of produce eaten falls too.

A one-off tray audit works for a single comparison. How to conduct a food waste audit walks through the logistics. If you want to watch an OVS or menu change play out over a whole semester, food waste tracking methods compared sets out the trade-offs between manual audits, scales and camera-based systems. Seneca Valley School District records plate waste at its tray return with CPIsight; the Seneca Valley case study shows what continuous tracking looks like in a working K-12 program, and our K-12 page covers how districts use the data.