Nausea with or without vomiting (NVP) is probably the most frequently reported medical complaint of pregnancy, but few studies have considered risk factors for its development. We used data from an ongoing epidemiological study of pregnancies in four regional centres. Mothers of infants with congenital malformations (n = 17,158) and a sample of normal infants (n = 5,329) were interviewed within 6 months of delivery by trained nurse-interviewers using a standardised questionnaire. For all risk factors investigated, odds ratios and 95% confidence intervals were calculated using multiple logistic regression, controlling for potential confounders. The cumulative incidence (risk) of NVP was 67%. The risk of NVP and its timing during pregnancy were similar for mothers of malformed and normal infants, so data were combined. No changes in the NVP risk were observed over the 20-year study period. The risk decreased with increasing age, but increased with increasing gravidity. The risk also increased with increasing number of prior miscarriages. Further, within each gravidity category, the risk was higher for twin births than for singletons. Women who reported onset of NVP after the first trimester differed demographically from women whose NVP began earlier: they were less-well educated, had lower incomes, and were more likely to be black. The finding that the number of prior pregnancies, both complete and incomplete, and number of fetuses independently appear to increase the risk of NVP suggests a fetal 'dose' effect. Together with selected demographic characteristics that differentiate early- vs. late-onset NVP, these findings warrant further investigation.