PIP: This study postulates that contraceptive use in Pakistan is affected by the usual demographic factors as well as husband-wife communication, female autonomy, son preference, religious beliefs, and family planning service supply. Analysis is based on data obtained from the Pakistan Demographic and Health Survey of 1990-91. Findings indicate that 74% of women never talked in the past year with their husbands about family planning. Almost 60% believed that family size was "up to God." About 47% knew where to obtain contraception; only 20.4% had easy access to a source of supplies. Current use was 14% and ever use was 22.4%. Analysis is based on three basic models. Model 1 includes the control variables and son preference. Model 2 includes husband-wife communication, religious attitudes, and female autonomy. Model 3 includes the addition of family planning to model 2 variables. Urban residence increases the odds of contraceptive use considerably only in Model 1. The influence of urban residence in the other models is reduced. Husband's education is significant only in Models 1 and 2 and insignificant in Model 3 when the family planning variable is included. Increased women's age is also insignificant in Model 3. Of the supply factors in Model 3, knowledge of a source and easy access to a source were highly significant, while mass media exposure was not important. Knowledge of a source was the most important predictor. Model 3 explained 90% of use. Among urban women, lack of husband-wife communication and fatalistic beliefs reduce the log-odds of contraceptive use. For rural women, age and women's secondary education were key predictors. Findings confirm that demographic and socio-cultural factors affect contraceptive use in Pakistan. All the theorized variables exerted a strong influence on contraceptive use, which can be counteracted by improved supply and service strategies.