Home garden interventions combining training in agriculture and nutrition have the potential to increase vegetable production and consumption in lower-income countries, but there remains a need for better evidence for impact. This study contributes to filling this gap by evaluating the impact and distributional effects of a home garden intervention in Cambodia. We used a cluster randomized controlled trial with before and after data for a sample of 500 rural households with children under five and women 16–49 years old. Impact was estimated using analysis of covariance (ANCOVA) and conditional quantile regressions with non-parametric bounds. The results show that the intervention significantly increased the adoption of nearly all promoted gardening methods. More households (+ 35%; p < 0.01) produced vegetables, and the production period was extended by five months on average. One-month recall data show an increase in vegetables harvested (+ 25 kg; p < 0.01) and consumed (+ 10 kg; p < 0.01) from the garden. Quantile regressions confirm these findings and show that nearly all households benefitted, but households that were already doing better at baseline tended to benefit more. Seven-day recall data show an increase in the quantity of vegetables consumed (+ 61 g/day/capita; p < 0.01) and an increased quantity of vitamin A, folate, iron, and zinc contained in these vegetables. Women contributed more to the garden work than men and their time spent in the garden increased by 29 min/day on average. These results add further evidence that integrated home garden interventions can contribute to nutrition outcomes and that almost all participants can benefit.