We conduct a field experiment in 31 primary schools in England to test the effectiveness of different temporary incentives on increasing choice and consumption of fruit and vegetables at lunchtime. In each treatment, pupils received a sticker for choosing a fruit or vegetable at lunch. They were eligible for an additional reward at the end of the week depending on the number of stickers accumulated, either individually (individual scheme) or in comparison to others (competition). Overall, we find no significant effect of the individual scheme, but positive effects of competition. For children who had margin to increase their consumption, competition increases choice of fruit and vegetables by 33% and consumption by 48%. These positive effects generally carry over to the week immediately following the treatment, but are not sustained effects six months later. We also find large differences in effectiveness across demographic characteristics such as age and gender.