This article analyzes the conceptual and methodological approaches which have been used to investigate effects of prenatal maternal stress on birth outcomes and highlights the major findings of this research. By viewing the most widely used operational definitions of prenatal stress in a broader theoretical framework, it can be seen that most studies have failed to conceptualize stress reliably. This, in addition to common methodological and design flaws which are described in the article, has produced equivocal findings about the role of stress in adverse birth outcomes such as preterm delivery and low birth weight. Recent studies using more powerful, multidimensional approaches to stress definition and measurement provide more definitive evidence and suggest some precise effects. Implications and strategies for future research are presented.