This study examined the determinants of blood lead (BPb) in 513 pregnant women in Mexico City: 311 from public hospital prenatal clinics, representing primarily women of low socioeconomic status (SES), and 202 from private hospitals, primarily women of high SES. Overall, BPb levels ranged from 1.38 to 29 micrograms/dl, with geometric means of 6.7 and 11.12 micrograms/dl for women from private and public hospitals, respectively. The crude geometric means difference obtained by t-test was 4.42 (p < 0.001). BPb was measured from January 1994 to August 1995 and showed higher levels during fall and winter and lower levels during spring and summer. The main BPb determinants were the use of lead-glazed ceramics in women from public hospitals and season of the year in women from private hospitals. Consumption of tortillas (corn bread rich in calcium) decreased BPb levels in the lower SES group, but the relationship was not statistically significant (p > 0.05). Consumption of milk products significantly (p < 0.05) reduced BPb levels in the higher SES group. In 112 women whose diets were deficient in calcium, taking calcium supplements lowered their blood lead levels about 7 micrograms/dl. A predictive model fitted to these data, using the strongest predictors plus gestational age, showed a difference of 14 micrograms/dl between the best and worst scenarios in women from public hospitals. Avoiding use of lead-glazed ceramics, consuming diets rich in calcium, and, if needed, taking calcium supplements, would be expected to result in substantial lowering of BPb, especially in pregnant women of low socioeconomic status.