Background: Food deserts are thought to be a barrier to making healthier food choices. This concept has been challenged. The interaction between the physical environment and children's food choice has received little attention. The present study used food intake data to generate hypotheses concerning the role of the physical environment in food choice.
Methods: A cross-sectional analysis was conducted of the dietary habits of Year 5 (9-10-year-old) children from 90 of Liverpool's 118 primary schools. Individuals with the 'best' and 'worst' food choices were mapped and two areas associated with these extreme choices located.
Results: One thousand five hundred and thirty-five children completed the dietary questionnaire and supplied a full and valid postcode. Two adjacent areas with relatively large numbers of children in the 'best' and 'worst' food choice groups were chosen. Both areas had very similar socio-economic profiles. The contrast in the physical environments was striking, even on visual inspection.
Conclusions: Food deserts as a cause of poor food choice did not stand scrutiny; the area located by the worst food choices had a plethora of shops selling food (better termed a food prairie), whereas the area located by the best food choices had no shops in evidence but did have more 'space'.