This paper estimates heterogeneous effects of food source (food away from home, at home and from school) on child dietary quality. Using a quantile estimator designed for panel data, two non-consecutive days of intake are used to identify the effiect of food source across the unconditional distribution of dietary quality. Main results suggest that food away from home has a negative impact on dietary quality for all children except those falling in the very lowest portion of the unconditional distribution. As compared to home-prepared food, school food is found to increase dietary quality for children falling in the bottom quartile of the distribution. For children with a very high underlying proneness to consume a healthful diet, food from school has a negative effect. While food consumed under the National School Lunch and Breakfast Programs may not benefit every child (especially at the mean), it does improve the diets of many children whom otherwise would have poorer dietary quality. The implication is that U.S. schools are fertile grounds to improve nutrition skill formation, especially for the most disadvantaged.