An experimental maternal and child health and family planning programme has been in existence in Matlab for almost 10 years. During this time the project has achieved remarkable success in the area of family planning. Based upon a 1984 survey, this study examines the pattern of contraceptive use in the Matlab treatment area, and contrasts it with the regular government programme in the neighbouring comparison area. Important differences between the two areas are observed, with the treatment area characterized by substantially higher levels of current contraceptive use, greater reliance upon temporary methods for birth spacing and, among acceptors of sterilization, more prior experimentation with other methods. In the comparison area lower rates of contraceptive use are observed, with heavy reliance upon tubectomy for limiting family size. The findings suggest that an intensive and innovative family planning programme in rural Bangladesh can achieve success not only in terms of contraceptive prevalence, but can also attract users interested in child spacing and others wanting to limit their family size, by offering the widest range of contraceptive methods.