Childhood obesity is prevalent, and dietary habits are a key determinant. Some children skip breakfast for weight control, but studies have shown mixed results. Therefore, we assessed the association between breakfast skipping and body mass index (BMI) among young Chinese children in Hong Kong. DESIGN/SUBJECTS AND METHODS: A cohort of 113,457 primary 4 participants of the Department of Health Student Health Service in 1998-2000 was followed up for 2 years, with 68,606 (60.5%) participants available for analysis in primary 6. The sociodemographic characteristics for traced and untraced participants were similar. At baseline and follow-up, students reported breakfast habit (consumed vs skipped) and other lifestyle characteristics using a standardized questionnaire. BMI was derived using height and weight measured by trained nurses. Multivariable linear regression was used to examine the cross-sectional relationship between breakfast habit and BMI, as well as the prospective association between baseline breakfast habit and change in BMI. Models adjusted for demographic, socioeconomic and lifestyle characteristics. A total of 1805 (5.3%) boys and 1793 (5.2%) girls skipped breakfast at baseline. In cross-sectional analyses, breakfast skippers had a higher mean BMI than did eaters among both primary 4 (β = 0.77, 95% confidence interval (95% CI): 0.67-0.87) (P < 0.001) and primary 6 children (β = 0.86, 95% CI: 0.78-0.95) (P < 0.001). Compared with eaters, baseline breakfast skippers experienced a greater increase in BMI in the subsequent 2 years (β = 0.11, 95% CI: 0.07-0.16) (P < 0.001), and this association was stronger among lunch skippers than eaters (P for interaction = 0.04). Our study provided prospective evidence that skipping breakfast predicts a greater increase in BMI among Hong Kong children. As breakfast is a modifiable dietary habit, our results may have important implications for weight control. However, the underlying mechanism of this effect warrants further investigation.