The impact of malaria on low birthweight was investigated in Bougoula village (Sikasso region, Mali). In two successive years, pregnant women were followed until delivery. Phase I (1992) was observational, with 135 complete observations. Phase II (1993) included 126 participants, who were offered malaria prophylaxis with proguanil (200 mg/day) and chloroquine (300 mg/week). The results show that 1) infants of first and second pregnancies had lower birth weights (-382.7 +/- 62.6 g; P < 0.0001) compared with higher rank pregnancies; 2) strong seasonal variation in birthweight was observed in Phase I, with an annual cycle, a nadir in January, and an amplitude of 372.4 g (P = 0.0002); 3) parasitemia measured during pregnancy was associated with lower birthweight in infants from first and second pregnancies, but not from higher parity mothers; and 4) malaria prophylaxis taken for 20 weeks or more in Phase II suppressed the seasonal variation of birthweight and the effect of low parity (+423.4 +/- 118.8 g; P = 0.0004). We conclude that malaria in pregnancy has an important negative impact on birthweight in first and second pregnancies. Prophylaxis with proguanil and chloroquine is an effective prophylaxis when taken for 20 weeks or more.