The contraceptive prevalence rate reported in the 1975 Sri Lanka Fertility Survey (SLFS) was lower than would have been expected, given the country's low level of fertility. Later surveys, such as the 1982 Contraceptive Prevalence Survey and the 1987 Demographic and Health Survey, showed a pattern of rising prevalence with little change in the level of fertility. In all three surveys, fertility levels predicted from reported contraceptive prevalence were higher than the actual rates, but the disparity was largest for 1975. The apparent discrepancy between fertility and contraceptive use was caused by substantial underreporting of use of traditional methods in the SLFS and more accurate reporting in successive surveys. Contraceptive use was understated primarily because traditional methods were not perceived by their users as methods of fertility control. Former users of traditional methods, who later accepted sterilization and were less reluctant to report its use, account for much of the observed increase in contraceptive prevalence between surveys.