Respiratory function test, arterial blood gas analysis, and ejection fraction were used to compare three protocols of breathing exercises during immersion in 38 degrees C water. Therapy was given for 2 mo to patients with stable chronic obstructive pulmonary disease. Protocol A consisted of a total exercise period of 20 min/wk (10 min/day, 2 days/wk) and was performed by 7 patients (5 cases of asthma and 2 cases of emphysema). Protocol B consisted of a total exercise period of 120 min/wk (20 min x 2 per day at 10:00 am and 3:00 pm, 3 days/wk) and was performed by 9 patients (6 asthmas and 3 emphysemas). Protocol C consisted of a total exercise period of 120 min/wk (20 min/day, 6 days/wk) and was performed by 8 patients (4 asthmas and 4 emphysemas). The ratio of forced expired volume in one second to forced vital capacity (FEV1.0%) was significantly increased in protocols B and C (P < 0.01). The ratio of forced vital capacity to the predicted normal value (%FVC) was not changed in any of the three protocols. A significant increase in peak flow was observed in protocols B and C (P < 0.05). The maximal expiratory flow at 25% (V25) was not changed in any of the three protocols. PaO2 was significantly increased and PacO2 was significantly decreased in protocol B (P < 0.01 and P < 0.05, respectively), whereas only PaCO2 was significantly decreased in protocol C (P < 0.05). Ejection fraction was increased in protocols B and C. These results suggest that exercise for a total period of 120 min/wk is preferable to that of 20 min/wk in COPD.