The aim of this study was to examine the relationship between feeding frequency (FFQ) and body mass index (BMI) in a free-living group of teenagers, since recent evidence suggests that this may be an artifact of under-reporting. The data was cross-sectional, and a sample of 731 respondents (M = 298, F = 433) from the 1970 Longitudinal Birth Cohort Study were investigated. An initial significant relationship between FFQ and BMI was found for both males and females. However, the removal of overweight males who were dieting and a group of non-dieting females with BMI < 25 who perceived themselves to be overweight negated this relationship for males and females respectively. In conclusion, the apparent relationship between FFQ and BMI observed in this and other free-living studies appears to be an artifact of dieting and dietary restraint rather than simply under-reporting.