An assumption made when designing recess interventions in schools is that there is a clear demarcation between eating time and play time. We used observational data conducted as part of the Sydney Playground Project to test if this assumption was correct. The Sydney Playground Project is a cluster randomised controlled trial of a recess intervention. The SPP was conducted in 12 Sydney schools and involved 214 children aged 5-7 years. Analysis of the baseline behavioural observations revealed an average of 18.5% of children’s free play time was spent eating. Twenty observations were used to explore the extent and nature of influence of eating on children’s play. For most children, eating reduced their capacity to engage fully in play. Some children were socially isolated and unable to play while eating. A small number of children seemed not to be affected by their own eating, but were affected by their peers’ eating. The results indicate that a substantial proportion of children’s free play time can be lost to eating. Although the methods used do not allow us to make causal inferences, the present findings suggest that eating had a negative impact on quality and quantity of play. Time spent eating should be considered when estimating time children have for free play. Consideration should also be given to reducing the amount of free play time that can inadvertently be spent eating.