Context:Forward propulsion in freestyle swimming is predominantly achieved through arm action. Few studies have assessed the effects of arm training on arm power and swimming performance, yet there have not been any investigations on the effects of arms-only swimming training on swimming performance and physiological responses to arm exercise.Purpose:To investigate the changes in arms-only and full-stroke swimming performance, movement economy and aerobic power after an arms-only swimming training program.Methods:Fifteen male county level swimmers were assigned either to an experimental (ES, n = 8) or control group (CS, n = 7). For six weeks ES performed arms-only freestyle swimming exercises for 20% of their weekly training distance three times per week, whereas CS performed their usual swimming training. Before and after the training program, both groups performed a) two time trials, 186 m using arms-only (186ARMS) and 372 m using full-stroke (372FULL) freestyle swimming, and b) an incremental arm-pulling exercise test. The time to complete the trials was recorded. Peak oxygen uptake (VO2peak), peak exercise intensity (EIpeak) submaximal oxygen uptake at 60 W (VO2−60) and exercise intensity at ventilatory threshold (VTW) were determined from the exercise test.Results:After training, ES had improved in 186ARMS (−14.2 ± 3.6%, P = .03), VO2−60 (−22.5 ± 2.3%, P = .04), EIpeak (+17.8 ± 4.2%, P = .03), and VTW (+18.9 ± 2.3%, P = .02), but not in VO2peak (P = .09) or in 372FULL (P = .07). None of the measures changed in CS (P > .05).Conclusion:Arms-only swimming training at 20% of the weekly training distance is an effective method to improve arm conditioning during the preparatory phase of the annual training cycle.