The purpose of our study was to evaluate the differences in exercise performance on the cycle and rowing ergometers associated with training in collegiate women rowers. We postulated that improvement in a nonspecific activity, such as cycling, would predominantly reflect training adaptations in the cardiovascular system, whereas physiological gains in rowing would represent combined cardiovascular and muscular adaptations. Seven members (mean±SD: age = 20±1 yr) of the women's varsity rowing team of Dartmouth College were tested at the start (November) and completion (May) of the training season. Incremental exercise testing was performed on successive days in random order on the cycle (Monark‐Crescent AB, Varberg, Sweden) and rowing (Concept II, Morrisville, VT) ergometers. After six months of training: 1) maximal oxygen consumption (VO2max) increased significantly for both cycling (+6 percent) and rowing (+12 percent); and 2) oxygen consumption (VO2) at the ventilatory threshold (VT) increased significantly for both cycling (+5 percent) and rowing (+23 percent). Comparison of the absolute changes in VO2maxand VO2(VT) showed that these physiological parameters increased significantly more for rowing than cycling(p<0.05). We conclude that the increments in VO2maxand VO2(VT) with training for competitive collegiate rowers reflect adaptations in both the cardiovascular and muscular systems. The greater physiological improvements with rowing compared to cycling presumably represent training‐induced changes in the specific muscles used in rowing. Comparative exercise testing may help to delineate possible specificity of training effects.