1. Recently, we have demonstrated that hypoxic breathing is followed by an increase in plasma digoxin-like substance in normal men. 2. This study was undertaken in order to evaluate whether or not a low arterial O-2 partial pressure is combined with an increase in plasma digoxin-like substance in chronic pathological conditions also. 3. Sixteen male patients (mean age 53.1+/-3.7 years) affected by chronic obstructive pulmonary disease of a mild stage were studied. They were further sub-divided according to their arterial O-2 partial pressure into 'mild hypox;ic' (n = 8, mean age 52.5+/-2.7 years), with an arterial O-2 partial pressure between 66 and 75 mmHg, and 'severe hypoxic' (n = 8, mean age 54.3+/-5.1 years), with an arterial O-2 partial pressure less than or equal to 65 mmHg, groups. Seven healthy men (mean age 48.5+/-4.8 years) voluntarily participated as the control group. 4. Plasma digoxin-like substance levels were significantly higher in 'severe hypoxic' patients (203.5+/-9.9 pg/ml) than in both 'mild hypoxic' patients (169.5+/-31.4 pg/ml, P<0.02) and normal subjects (158.9+/-12.4 pg/ml, P<0.0001) and were directly correlated with urinary Na+ excretion (severe hypoxic group, r = 0.756, P<0.007; mild hypoxic group, r = 0.789, P<0.02). Considering the two hypoxic groups together, plasma digoxin-like substance levels were negatively correlated with arterial O-2 partial pressure (r = -0.740, P<0.001). However, when patients were subdivided according to the degree of hypoxia, the negative correlation between arterial O-2 partial pressure and plasma digoxin-like substance persisted only in severe hypoxic patients (r = -0.761, P<0.03). 5. These results indicate a clear linkage between blood O-2 and endogenous digoxin-like substance regulation. In particular, hypoxia may be regarded as an important stimulus for the release of the endogenous digoxin-like substance in man.