The purpose of this study was to determine whether trained competitive runners could maintain on-land running performance using 4 wk of deep water run training instead of on-land training. Eleven well-trained competitive runners (10 males, 1 female; ages, 32.5 +/- 5.4 yr; height, 179.8 +/- 9.3 cm; weight, 70.4 +/- 6.7 kg (mean +/- SD)) trained exclusively using deep water run training for 4 wk. Subjects trained 5-6 d.wk-1 for a total of 20-24 sessions (mean +/- SD, 22 +/- 1.5 sessions). Instruction and practice sessions were conducted prior to the training period. Before and after the deep water run training, subjects completed a 5-km race on the treadmill using a computer based system, a submaximal run at the same absolute workload to assess running economy, and a combined lactate threshold and maximal oxygen consumption test. No significant differences were found for (mean +/- SEM): 5-km run time (pre, 1142.7 +/- 39.5 s; post, 1149.8 +/- 36.9 s; P = 0.28), submaximal oxygen consumption (pre 44.8 +/- 1.2 mL.kg-1.min-1; post, 45.3 +/- 1.5 mL.kg-1.min-1; P = 0.47), lactate threshold running velocity (pre, 249.1 +/- 0.9 m.min-1; post, 253.6 +/- 6.3 m.min-1; P = 0.44), or maximal oxygen consumption (pre, 63.4 +/- 1.3 mL.kg-1.min-1; post, 62.2 +/- 1.3 mL.kg-1.min-1; P = 0.11). Also no differences were found among Global Mood State pre-training, each week during training, and post-training. Competitive distance runners maintained running performance using 4 wk of deep water run training as a replacement for on-land training.