This study examined the interactive effects of pregnancy and aerobic conditioning on maternal cardiac structure and function. Effects of closely monitored cycle ergometer conditioning were studied during the second (TM2) and third trimesters (TM3) in 22 previously sedentary pregnant women (exercised group, EG) and a nonexercising pregnant control group with similar characteristics (CG, n = 19). Subjects were studied in the resting state by two-dimensional echocardiography and during cycle ergometer exercise at three steady-state power outputs at the start of TM2 (ENTRY), at the end of TM2 and TM3 (postconditioning), and 3-4 months postpartum (NPR, nonpregnant reference, CG only). Aerobic conditioning did not increase left ventricular dimensions beyond those attributable to pregnancy itself. In addition, in contrast with previous studies of nonpregnant women, physical conditioning during pregnancy did not reduce heart rate (HR) in the resting state. During exercise, the slope of the HR versus oxygen uptake (VO2) regression decreased significantly between preconditioning and the end of TM3 in the EG, suggesting that training-induced reductions in HR become more evident with increasing exercise intensity. Also, significant reductions in oxygen pulse (VO2/HR) were observed at all three work rates in the CG, but not in the EG. These findings support the hypothesis that the cardiovascular effects of aerobic conditioning are obscured by more powerful effects of pregnancy in the resting state but become "unmasked" during strenuous exercise.