This study evaluates risk factors associated with low birthweight in an African-American population. Records of 225 women delivering liveborn, nonanomalous singletons weighing < 2500 g were reviewed. The next parturient, matched for race only, of a similar infant weighing > or = 2500 g constituted the control. This case-control study was conducted among women delivering at University Hospital in New Orleans during 1996-1997. Mothers of infants weighing < 2500 g were more likely to not have finished high school (49% versus 38%), to have received no prenatal care (26% versus 7%), or to have five or fewer visits if care was obtained (52% versus 33%). The mother was more likely to weigh < 60 kg (49% versus 32%), to smoke (24% versus 11%), or to have used cocaine (18% versus 5%) or alcohol (11% versus 5%). Parturients of low birthweight newborns were more likely to have had a prior low birthweight infant (44% versus 19%) and themselves to have had a birthweight < 2500 g (30% versus 13%). Regression analysis confirmed the importance of three parameters as associated with low birthweight: no prenatal care (odds ratio [OR] = 6.0 [1.1-31.4]), alcohol use (OR = 5.2 [1.1-24.8], and low maternal birthweight (OR = 3.9 [1.9-7.9]. These results indicate that evaluations of low birthweight in African Americans should consider maternal birthweight and that efforts to improve pregnancy outcome should be structured in terms of generations.