The effect of protriptyline, a tricyclic antidepressant, on sleep architecture, nocturnal arterial oxygen desaturation, pulmonary function, and diurnal arterial blood gases was investigated in an open study of 14 patients with stable chronic obstructive lung disease. Daytime and overnight measurements were made before and 2 and 10 weeks after they started protriptyline (20 mg daily at bedtime). Two patients had to be excluded before the second visit and one before the third visit because of changes in treatment for their chest disease. Protriptyline caused mouth dryness in all patients and dysuria in six men. With protriptyline there were no significant changes in total sleep time, sleep period time, or the percentages of total sleep time occupied by stage I-II and stage III-IV sleep. The mean (SEM) percentage of total sleep time spent in rapid eye movement (REM) sleep decreased from 11.1 (1.7) to 4.6 (0.7) at two weeks and to 4.2 (1.0) at 10 weeks. After protriptyline the time spent during sleep with an arterial oxygen saturation (SaO2) below each 5% increment above 65% was less than the baseline time; the lowest SaO2 (%) reached during sleep increased from 64.5 (1.7) to 72.7 (2.1) at 2 weeks and to 77.4 (2.1) at 10 weeks. Lung volumes and expiratory flows were unchanged during the study. Daytime arterial oxygen tension (PaO2) increased from 57 (1.4) mm Hg before treatment to 62 (1.9) mm Hg at 2 weeks and to 66 (1.9) mm Hg at 10 weeks (7.6 (0.2), 8.3 (0.3), 8.8 (0.3) kPa). Carbon dioxide tension fell from 52 (2.3) mm Hg to 49 (1.4) mm Hg at 2 weeks and to 48 (2.0) mm Hg at 10 weeks (6.9 (0.3), 6.5 (0.2), 6.4 (0.3) kPa), but these changes were not significant. These results suggest that protriptyline may benefit patients with chronic obstructive lung disease by reducing the sleep induced falls in SaO2 and improving diurnal PaO2; a controlled trial is now required.