Objective: To assess whether a Smarter Lunchroom intervention based on behavioural economics and adapted for students with intellectual and developmental disabilities would increase the selection and consumption of fruits, vegetables and whole grains, and reduce the selection and consumption of refined grains. Design: The 3-month intervention took place at a residential school between March and June 2012. The evaluation employed a quasi-experimental, pre-post design comparing five matched days of dietary data. Selection and plate waste of foods at lunch were assessed using digital photography. Consumption was estimated from plate waste. Setting: Massachusetts, USA. Subjects: Students (n 43) aged 11-22 years with intellectual and developmental disabilities attending a residential school. Results: Daily selection of whole grains increased by a mean of 0?44 servings (baseline 1?62 servings, P 5 0?005) and refined grains decreased by a mean of 0?33 servings (baseline 0?82 servings, P 5 0?005). The daily consumption of fruits increased by a mean of 0?18 servings (baseline 0?39 servings, P 5 0?008), whole grains increased by 0?38 servings (baseline 1?44 servings, P 5 0?008) and refined grains decreased by a mean of 0?31 servings (baseline 0?68 servings, P 5 0?004). Total kilojoules and total gram weight of food selected and consumed were unchanged. Fruit (P 5 0?04) and vegetable (P 5 0?03) plate waste decreased. Conclusions: A Smarter Lunchroom intervention significantly increased whole grain selection and consumption, reduced refined grain selection and consumption, increased fruit consumption, and reduced fruit and vegetable plate waste. Nudge approaches may be effective for improving the food selection and consumption habits of adolescents and young adults with intellectual and developmental disabilities.