OBJECTIVE: This study had two aims: (1) to examine pregnant women's alcohol consumption across time from prepregnancy until childbirth and (2) to explore whether prepregnancy drinking and intention to drink predict prenatal alcohol consumption while controlling for relevant demographic variables. METHODS: At 17-21 weeks, 248 pregnant women completed questions about demographics, intention to drink alcohol during the subsequent pregnancy, and retrospective measures of prepregnancy and early pregnancy consumption. After this time, calendars were sent fortnightly assessing daily alcohol consumption until birth. RESULTS: For women who drank both prepregnancy and postpregnancy confirmation, average fortnight alcohol consumption in the first weeks of pregnancy was lower than during prepregnancy, and consumption continued to decrease between gestational weeks 1 and 8, particularly following pregnancy confirmation, after which it remained relatively stable. When predicting whether women drank in late pregnancy, intention accounted for unique variance after controlling for income and prepregnancy drinking. For women who drank after pregnancy confirmation, prepregnancy drinking quantity significantly predicted intention to drink, which in turn predicted fortnight alcohol consumption in later pregnancy, after controlling for prepregnancy drinking and income. CONCLUSIONS: Findings highlight the need to measure alcohol consumption at multiple time points across pregnancy, the need for educating and supporting women to reduce consumption when planning pregnancies, and the usefulness of intention to drink as a predictor of drinking during pregnancy.