The objective of this study was to measure the independent effects of clinical factors and non-clinical factors, such as provider and sociodemographic characteristics, on the number of antenatal visits made by women in England and Wales. The study was based on a survey of the secondary case records of 20,771 women with singleton pregnancies who were delivered between 1 August 1994 and 31 July 1995. The women in the survey attended one of nine maternity units in Northern England and North Wales selected within those areas to reflect geographical variations, as well as variations in the size and teaching status of the institution. A multivariate Poisson regression model was developed to examine differences in the number of antenatal visits made by women with different clinical and non-clinical characteristics. After controlling for non-clinical factors, primiparous women identified as high risk at booking made 1.0% more visits than primiparous women identified as low risk at booking (p = 0.196). Multiparous women identified as high risk at booking made 3.5% more visits than their low risk counterparts (p<0.001). High risk-defining criteria during antenatal care led to a 0.3% weekly increase in the number of antenatal visits amongst primiparous women (p <0.001) and a 0.4% weekly increase in the number of antenatal visits amongst multiparous women (p < 0.001). Several notable results, not reported elsewhere in the literature, were revealed by the regression analyses. After all independent variables were controlled for, women who were booked into urban teaching hospitals made 10% fewer antenatal visits than the women who were booked into the urban non-teaching hospitals. Women of Pakistani origin made 9.1% fewer antenatal visits than women of white British origin. Similar results were revealed for women of Indian origin and women from other ethnic groups. Non-smokers made 6.0% more antenatal visits than smokers. The planned pattern of antenatal care, number of carers seen, gestation at first presentation and maternal age also had significant independent impacts on the number of antenatal visits. The study highlights the sizeable impact of non-clinical factors on the antenatal care delivery process and indicates ways in which variations in antenatal care might be reduced.