Objectives To determine (a) whether the influence of the determinants of family planning use in Pakistan changed between 1990–91 and 2006–07, and (b) if these changes were associated with changes in the method mix.

Methods Data from the Pakistan Demographic and Health Surveys (PDHS) of 1990–91 and 2006–07 were used in the analyses. Data on 5184 married, non-pregnant, fecund women in 1990–91 and 8041 married, non-pregnant, fecund women in 2006–07 were used. Logistic regression analysis was used to identify factors associated with the use of any contraceptive method and whether the influence of these factors changed between the survey years. Changes in the method mix were examined.

Results The effects of urban/rural residence, wealth and education on contraceptive use changed between 1990–91 and 2006–07. Differentials in contraceptive use by residence, wealth and education declined and were accompanied by changes in the method mix. In rural areas and among less-educated women, the contribution of traditional methods to the method mix increased. Among the poorest women, the method mix shifted towards traditional methods and condoms.

Conclusions Less-educated, rural, Pakistani women increased the use of family planning at a faster rate than more-educated, urban, women by adopting the use of traditional family planning methods. Poor women also increased family planning use more quickly than non-poor women, by adopting condoms and traditional methods. The more rapid increase in the demand for family planning among poorer, less-educated, rural women is a positive trend. In order to convert this demand into the use of longer-term modern methods, however, access to high quality services must be improved in rural and low-income urban areas.