The study purpose was to contribute to the development of a model for preterm birth that brings together social, psychosocial and biomedical factors. The three research questions were: (a) Which social and psychosocial factors influence the risk of preterm delivery? (b) Are the effects of social and psychosocial factors independent? (c) Do the biomedical factors identified and measured here explain the effect of social and psychosocial factors? The sample comprised 739 low-income black non-Hispanic women. Interviews were conducted after delivery, and medical records were abstracted. Nearly a quarter of the women delivered preterm (23.5%). Two social factors strongly predicted risk of preterm delivery: inadequacy of time and money for non-essentials. The effect of both was stronger than expected from each alone and was partially mediated by psychosocial factors. Several psychosocial factors also predicted preterm delivery risk but only stress and locus of control were independent predictors. With the inclusion of biomedical factors, stress (OR = 1.86, P = 0.005) and locus of control (OR = 1.75, P = 0.007) continued to be strongly associated with preterm delivery. The effect of inadequate resources for non-essentials was no longer significant, suggesting mediation. These strong effects of social and psychosocial factors should be examined in future studies.