Background While it appears that social capital has a positive effect on mental health, most studies have been cross-sectional in nature and/or employ weak measures of social capital or mental health. Even less attention has been paid to vulnerable populations, such as lowincome women with children. Thus, our objective was to explore how different dimensions of social capital impact depression in this population. Methods We used data from the Fragile Families and Child Wellbeing Study, which has followed a cohort of children born in large US cities to mostly minority, unmarried parents for over 9 years. These families tend to be at greater risk for falling into poverty. Four separate measures of social capital were constructed, using measures that are reliable and that offer evidence of validity including social support and trust, social participation, perceptions of neighbourhood social cohesion and perceptions of neighbourhood social control. The temporal effect of social capital on mental health, as measured by a standardised screening for depression was investigated using logistic regression. Results After controlling for relevant socioeconomic and demographic factors, prior depression, and prior self-rated health, the social capital dimensions of social support and trust and perceived neighbourhood social cohesion are significant predictors of depression. Conclusions These results suggest that social and neighbourhood environments play an important role in mental health status. Intervention and policy initiatives that increase social capital may be viable for improving mental health among low-income urban, minority women.