BACKGROUND: Epidemiological data on obesity prevalence are scarce in Romanian population. Consequently, the aim of our study was to evaluate the prevalence of obesity and unhealthy behaviors among school children and adolescents from Bucharest, Romania. METHODS: Cross-sectional study, 866 participants (53.2% girls, 46.8% boys, age range 6-18 years), selected by systematic sampling with probability-proportionate-to-size from all Bucharest's schools. MEASUREMENTS: height, weight and a questionnaire to collect information about life style and eating behavior. Nutritional status was established based on World Health Organization recommendations (WHO), International Obesity Task Force (IOTF), Center for Diseases Control (USA-CDC) cut off values and local standards, respective. RESULTS: The prevalence of overweight (including obese) and obesity alone based on different standards, was 31.6% and 11.4% (WHO), 24.6% and 6.2% respectively (IOTF), 25.2% and 10% (USA-CDC), 22.3% and 12.5% (local standards). When using local standards (weight only based), the obese subjects proportion among overweight children raised questions regarding the appropriateness of these standards. Overweight (including obese) prevalence was significantly higher among the boys versus girls: 36.2% vs. 27.6%, ( OR 1.5; 95% CI 1.12-2.03; p value = 0.006) and among the 6-10.9 years vs. 11-17.9 age group, (40.7% vs 26.6%). Almost all the participants (95%) reported at list one unhealthy eating behavior but no significant relationship was found with overweight or obesity only. CONCLUSIONS: This first epidemiological study of obesity prevalence in school children and adolescents showed that 11.4% of Bucharest's children and adolescents were obese by WHO classification, 6.1% by IOTF cut off values and 10% by CDC classification. Younger children and the boys were more affected no matter which standard we used. In spite of unsignificant relationship to the adiposity status, our data showed a high prevalence of unhealthy eating behaviors reported by the participants. Particular aspects of the overweight versus obesity prevalence, after applying local standards, suggests that international recognized algorithms should be used for constant epidemiological evaluation instead of establishing local criteria.