Background. Alcohol-related mortality and morbidity among women has increased over recent decades, especially in areas of higher deprivation. Pre-pregnancy alcohol use is associated with continued consumption in pregnancy. We assessed whether general population alcohol consumption patterns were reflected among pregnant women in two Scottish areas with different deprivation levels. Methods. Cross-sectional study in two health boards (HB1, lower deprivation levels, n=274; HB2, higher deprivation levels, n=236) using face-to-face 7-day Retrospective Diary estimation of peri-conceptual and mid-pregnancy alcohol consumption. Results. A greater proportion of women in HB2 (higher deprivation area) sometimes drank peri-conceptually, but women in HB1 (lower deprivation area) were more likely to drink every week (49.6% vs 29.7%; p<.001) and to exceed daily limits (six units) at least once each week (32.1% vs 14.8%; p<.001). Following pregnancy recognition consumption levels fell sharply, but women in HB2 were more likely to drink above recommended daily limits (two units) each week (2.5% vs 0%; p<.05). However, women in HB1 were more likely to drink frequently. Women with the highest deprivation scores in each area drank on average less than women with the lowest deprivation scores. Conclusions. Heavy episodic and frequent consumption was more common in the lower deprivation area, in contrast with general population data. Eliciting a detailed alcohol history at the antenatal booking visit, and not simply establishing whether the woman is currently drinking, is essential. Inconsistent messages about the effects of alcohol in pregnancy may have contributed to the mixed picture we found concerning peri-conceptual and mid-pregnancy alcohol consumption.