Background and methodology This study aimed to investigate associations between area-level socioeconomic disadvantage (central heating, car ownership and residents in professional occupations), individual-level socioeconomic position (social class and educational qualifications) and contraception use in the UK for the period 1990–1991. Multilevel logistic regression analysis was conducted on cross-sectional data from the National Survey of Attitudes and Lifestyles of 9793 women, 16–59 years of age, residing in 646 postcode districts throughout the UK.

Results Women with lower levels of formal education were less likely to use contraception than women with higher education [odds ratio (OR) 0.50, 95% CI 0.44–0.57]. Women in the middle and low social class groups were less likely to use contraception than women in the higher social class group (OR 0.84, 95% CI 0.74–0.97 and OR 0.66, 95% CI 0.56–0.79, respectively). The association between social class and contraception use varied significantly across postcode districts (p<0.001). The contraception use of women in the lowest social class group varied the most geographically. Women in the lowest quintiles of disadvantage were less likely to use contraception than women in the most advantaged quintiles according to all three measures, namely central heating (OR 0.76, 95% CI 0.61–0.94), car ownership (OR 0.67, 95% CI 0.53–0.84) and residents in professional occupations (OR 0.75, 95% CI 0.61–0.93).

Discussion and conclusion Although more information is needed to understand how area and individual socioeconomic characteristics are associated with contraceptive use, this study suggests that policy on contraceptive use needs to be extended beyond individually targeted approaches and needs to take into account socioeconomic determinants of contraceptive use.