The purpose of this study was to compare the effects of 10 weeks of effort-matched short intervals (SI; n = 9) or long intervals (LI; n = 7) in cyclists. The high-intensity interval sessions (HIT) were performed twice a week interspersed with low-intensity training. There were no differences between groups at pretest. There were no differences between groups in total volume of both HIT and low-intensity training. The SI group achieved a larger relative improvement in VO(2max) than the LI group (8.7% ± 5.0% vs 2.6% ± 5.2%), respectively, P ≤ 0.05). Mean effect size (ES) of the relative improvement in all measured parameters, including performance measured as mean power output during 30-s all-out, 5-min all-out, and 40-min all-out tests revealed a moderate-to-large effect of SI training vs LI training (ES range was 0.86-1.54). These results suggest that the present SI protocol induces superior training adaptations on both the high-power region and lower power region of cyclists' power profile compared with the present LI protocol.