This methodological paper elaborates on the pitfalls and potentials for using participatory photo interviews that explore children's experiences with food. It aims to show how participatory photo interviews can give a deeper understanding of children's food preferences. Such knowledge will be useful to other researchers who wish to apply the method in social studies involving children and food. We studied 12 families comprising seven‐ and eight‐year‐old children and their parents to illustrate how this method can be used to identify and understand children's food preferences. The children took photographs over a one‐week period while eating family dinners at home and grocery shopping with their parents. The photos were then used to elicit information during separate in‐depth interviews. Qualitative data analysis software (NVivo 10) was used to content analyse the data. We found that participant photo interviews were a useful method for helping children to remember and describe the sensory, cognitive and affective situational associations of their meals. This approach helped to explore interesting aspects of children's preferences such as their response to raw vegetables, the importance of controlling and choosing meal ingredients and ambivalent food preferences. We discuss parental involvement during data collection as it represents one of the methodological pitfalls. Another limitation is the risk of sensitive information to be breached. We conclude that participatory photo interviews have potential for future food research involving children and may provide information difficult to elicit using other methods.