This study examined associations between sleep onset latency and emotional eating in a minority sample of children. A cross-sectional school-based study of sleep, psychological constructs, diet and physical activity was conducted in 6 public and private schools in Los Angeles County. An ethnically diverse sample of 356 third through fifth graders completed confidential self-report surveys. Multilevel regression (MLM) analyses were conducted to study associations while controlling for gender, ethnicity, and the random effect of school. Girls made up 57% of the total sample, which was predominantly Latino (42.6%), followed by African Americans (21.6%) and Asians (19.2%). MLM revealed that there were significant associations between sleep onset latency and emotional eating (p=.030), depressive symptomology (p<.0001) and trait anxiety (p<.0001). Sobel's test for mediation showed that trait anxiety (p=.011) but not depressive symptomology (p=.141) was a mediator of the relationship between sleep onset latency and emotional eating. Thereby providing a mechanism through which sleep onset latency is related to emotional eating. These findings suggest that sleep onset latency is associated with increased anxiety, depressive symptoms, and emotional eating. Although causal inferences cannot be drawn from this cross-sectional data, future studies should examine the possibility that problems falling asleep could lead to emotional dysregulation that in turn leads to emotional eating. Emotional eating may be one avenue by which sleep disturbances lead to overweight and obesity.