BACKGROUND: The aim of this retrospective cohort study was to record the prevalence of overweight and obesity in relation to parental education level, parental body mass index and region of residence, in preschool children in Greece. METHODS: A total of 2374 children (1218 males and 1156 females) aged 1-5 years, stratified by parental educational level (Census 1999), were examined from 105 nurseries in five counties, from April 2003 to July 2004, Weight (kg) and height (cm) were obtained and BMI (kg/m2) was calculated. Both the US Centers for Disease Control (CDC) and the International Obesity Task Force (IOTF) methods were used to classify each child as "normal", "at risk of overweight" and "overweight". Parental demographic characteristics, such as age and educational level and parental anthropometrical data, such as stature and body weight, were also recorded with the use of a specifically designed questionnaire. RESULTS: The overall estimates of at risk of overweight and overweight using the CDC method was 31.9%, 10.6 percentage points higher than the IOTF estimate of 21.3% and this difference was significant (p < 0.001). Children with one obese parent had 91% greater odds for being overweight compared to those with no obese parent, while the likelihood for being overweight was 2.38 times greater for children with two obese parents in the multivariate model. CONCLUSION: Both methods used to assess prevalence of obesity have demonstrated that a high percentage of the preschool children in our sample were overweight. Parental body mass index was also shown to be an obesity risk factor in very young children.