This study was designed to test the hypothesis that the amount of oxygen required to complete a specific treadmill exercise protocol changes during pregnancy as a function of gestational age, weight gain, and an individual's exercise regimen. The amount of oxygen required to complete a three-step, graded workload treadmill protocol was monitored before, every 6 to 8 weeks during, and after a clinically normal, accurately dated, singleton pregnancy in 18 fit, recreational athletes. Nine subjects continued a moderate- to high-intensity exercise regimen and nine did not. The amount of oxygen required to complete each stage of the protocol decreased significantly (6% to 15%) in both groups in early pregnancy. In the women who continued exercise, the amount of oxygen required remained at or significantly below that required before conception for the remainder of the pregnancy and the initial postpartum period. Conversely, in the women who discontinued regular exercise, the amount of oxygen required for each stage rose progressively throughout the remainder of pregnancy at an average rate of 2% per lunar month. It peaked in the thirty-seventh week and remained significantly elevated 7 weeks post partum. However, when the data were corrected for the additional requirement imposed by pregnancy weight gain, net efficiency was increased in both groups throughout pregnancy. These data support three conclusions. First, the efficacy of low- to moderate-intensity treadmill exercise in physically active women is improved in early pregnancy. Second, this increased efficiency is masked later in pregnancy by the effects of weight gain. Third, the degree of efficiency is enhanced throughout pregnancy in women who continue a regular exercise regimen at or above a basic conditioning level.