The authors analyzed data from the Coronary Artery Risk Development in Young Adults (CARDIA) Study in order to examine associations between parity and lipoproteins. Of 2,787 women recruited in 1985-1986, 2,534 (91%) returned in 1987-1988 and 2,393 (86%) returned in 1990-1991 for repeat evaluations. Two-year change (1987-1988 to 1985-1986) in high density lipoprotein (HDL) cholesterol was significantly different among the parity groups. HDL cholesterol decreased in women who had their first pregnancy of at least 28 weeks duration during follow-up (mean +/- standard error, -3.5 +/- 1.2 mg/dl), and this change was significantly different from the increase in women parous at baseline who had no further pregnancies (2.5 +/- 0.3 mg/dl) and in nullipara (2.4 +/- 0.3 mg/dl). There was a nonsignificant trend for a greater decrease in HDL2 cholesterol fraction in the primipara compared with the other groups. The HDL cholesterol decrease remained significant after controlling for race, age, education, oral contraceptive use, and changes in body mass index, waist-hip ratio, physical activity, smoking status, and alcohol intake. Change in HDL cholesterol was also significantly different among the parity groups in analyses of pregnancies that occurred during the subsequent 3 years of follow-up. There were no differences for change in LDL cholesterol or triglycerides. Potential mechanisms for a detrimental effect of pregnancy on HDL cholesterol include hormonal, body composition, or life-style/behavioral changes.