In large, prospective studies of pregnancy conducted in the 1960s, women reported very accurately whether or not they smoked. However, in the 1990s, pregnant women who smoke are often pressured to reduce or quit smoking, and the incentive to misreport may be greater than in the past. To assess the accuracy of reported smoking, the authors compared self-reported smoking with cotinine in the serum and/or urine of 105 women who participated in the Calcium for Pre-eclampsia Prevention pilot study in 1992. Cotinine confirmed the report of 84.6% of women who reported smoking and 94.5% of women who denied smoking. These fractions are virtually identical to those obtained in a pregnancy cohort from the 1960s. The authors conclude that in the setting of two obstetrical research studies not specifically focused on smoking, the accuracy of self-reported cigarette smoking did not change substantially from the 1960s to the 1990s.