Progestins stimulate respiration. We have previously shown prolonged ventilatory improvement in chronic respiratory failure with short-term medroxyprogesterone acetate (MPA). The mechanism of the sustained respiratory effect is unknown. Insulin-like growth factor-I (IGF-I) and insulin have anabolic effects which could also improve ventilation in the long-term. To better understand the interactions between hormones and control of breathing, we evaluated the degree and duration of changes in IGF-I, insulin and cortisol after short-term MPA therapy in chronic respiratory insufficiency. Fourteen postmenopausal women with permanent or episodic hypercapnic or hypoxaemic respiratory failure were recruited for a placebo-controlled single-blind trial. After 14 days of placebo treatment and 7 days of washout, a daily dose of 60 mg MPA was administered for 14 days. Serum IGF-I, insulin and cortisol were measured five times at 3-week intervals: at baseline, after 14 days on placebo, after 14 days on MPA, and during the washout, on days 21 and 42. Serum IGF-I levels were 15.2 (SD 4.6), 20.8 (SD 6.8) and 17.2 (SD 6.4) at baseline, on MPA and after a 3-week washout. Serum insulin levels did not change [12.5 mU l(-1) (SD 4.1), 12.2 mU l(-1) (SD 4.8) and 14.5 mU l(-1) (SD 3.6), respectively]. Serum cortisol did not change. On MPA, IGF-I increased on average by 5.6 nmol l(-1) [95% confidence interval (95% CI) 1.4 to 9.9] or 42.0% (95% CI 6.3 to 77.8) from baseline. The IGF-I response coincided with the previously reported ventilatory improvement. MPA 60 mg daily for 2 weeks increases serum IGF-I in postmenopausal women with chronic respiratory insufficiency. During follow-up after MPA, there was a trend towards increased IGF-I and insulin levels. The role of these two hormones to induce prolonged ventilatory stimulation could not be excluded and further studies in larger populations are warranted.