Objective: To assess the concentration of serum leptin in a population of malnourished children and to compare the leptin levels of disabled and non-disabled children in this population. Design: Case -control study. Subjects: Eighty-one children, comprising 41 children with mixed disabilities and 40 non-disabled controls, were selected from 425 children involved in a case -control study assessing the nutritional status of children with disabilities in an Indian slum population. Methods: Leptin was measured in the serum samples and was compared with anthropometry (weight-for-age Z-scores (WAZ), height-for-age Z-scores (HAZ), weight-for-height Z-scores (WHZ), body mass index (BMI), mid-upper-arm circumference (MUAC), sub-scapular skinfold thickness and triceps skinfold thickness) and serum acute phase proteins. Results: The children were very malnourished with WAZ ¼ 7 2.07 (s.d. 1.15), HAZ ¼ 7 2.15 (s.d. 1.85) and WHZ ¼ 7 1.07 (s.d. 0.83). Leptin was extremely low in both the disabled (1.44 ng=ml; 95% confidence interval, CI, 1.23 -1.69) and the nondisabled (1.19 ng=ml; 95% CI 1.04 -1.35) children. There were no differences between the disabled and non-disabled groups as a whole but 15 children with neurological disabilities had significantly higher (P < 0.05) serum leptin (1.65 ng=ml; 95% CI 1.29 -2.06) than the non-disabled children. Girls (1.55 ng=ml; 95% CI 1.29 -1.87) had significantly higher concentrations of leptin than boys (1.11 ng=ml; 95% CI 1.02 -1.22; P ¼ 0.002). Leptin did not correlate with any biochemical or anthropometric measures. Conclusions: In this population, where malnutrition was common, serum leptin levels were very low and did not correlate with anthropometry.