Background Fetal undernutrition (of which low birth weight is a marker) is implicated in the development of type 2 diabetes. However, there is limited information on the extent to which birth weight is related to precursors of type 2 diabetes in childhood, both in individuals and between ethnic groups. We examined the associations between birth weight and precursors of type 2 diabetes in 9–10 year old children and the contribution of ethnic differences in birth weight to ethnic differences in diabetes precursors. Methods This investigation was based on the Child Heart and Health Study in England, a cross-sectional study of anthropometric and fasting metabolic markers (insulin, glucose, HbA1c, urate, C-reactive protein and lipids) in 9–10 year-old children predominantly of South Asian, black African-Caribbean and white European origin. Birth weight was obtained from a combination of birth records, Office for National Statistics data, maternity databases and parental recall. Associations between cardiometabolic risk markers and birth weight were estimated using multilevel linear regression adjusted for age, sex, ethnic group and school (fitted as a random effect). Results Birth weight was available for 3744 of 5004 (75%) study participants. Birth weight was inversely associated with urate in unadjusted analyses. After adjustment for height, lower birth weight was associated with higher fasting insulin (4.2% [95% CI 0.9%, 7.4%,]), HbA1c (0.4% [95% CI 0.1%, 0.8%]), fasting glucose (0.6% [95% CI 0.2%, 1.1%,]), urate (5.2% [95% CI 3.9%, 6.6%,]) and triglyceride (3.1 [95% CI 0.9%, 5.2%]) per 1 kg lower birth weight but was not associated with HDL-cholesterol or C-reactive protein. Associations were little affected by adjustment for gestational age. Birth weight was lower among children of South Asian and (to a lesser extent) black African-Caribbean origin by 231g (95% CI 183g, 280g) and 81g (95% CI 30g, 132g) respectively. However, adjustment for birth weight had little or no effect on the higher levels of insulin, HbA1c, C-reactive protein and triglyceride among South Asians when compared with white Europeans. Similarly, adjustment for birth weight had little or no effect on the higher levels of insulin and HbA1c among black African-Caribbeans compared with white Europeans. Conclusion Birth weight was inversely associated with risk markers for type 2 diabetes in childhood, but only after adjustment for height. However, lower birth weights in South Asians and black African-Caribbeans did not explain ethnic differences in risk markers for type 2 diabetes in childhood.