We compared the value of exercise testing and measurement of pulmonary haemodynamics (PH) in the pre-operative assessment of 5 patients (mean age: 64 years, 3 men) with clinical stage I or II bronchogenic carcinoma and severe chronic obstructive pulmonary disease. They were considered at high risk due to poor pulmonary function tests (PFT); (one or more of the following): (1) radionuclide calculated postlobectomy FEV1 < 30% predicted, (2) diffusion capacity or transfer factor < 60% predicted, combined with a fall in PaO2 on maximal exercise of > 5 mm Hg, (3) a PaCO2 at rest of > 45 mm Hg. Maximal oxygen uptake (VO2max) during symptom-limited cycle ergometry and PH were measured in these 5 patients. They were considered eligible for lobectomy if they fulfilled at least one of the two criteria: (1) mean pulmonary artery pressure (PAP) of < 35 mm Hg and pulmonary vascular resistance of < 190 dyn.s.cm-5 at moderate exercise (40 W), (2) a VO2max of > or = 15 ml/kg/min. Six months postoperatively PFT and VO2max were measured again. PAP40W was 21, 38, 38, 46 and 52 mm Hg, respectively, which would have excluded 4/5 patients from surgery. VO2max was 21.7, 14.9, 13.4, 19.2 and 18.6 ml/kg/min, respectively, which would have excluded 2/5 patients. Expressed in percent predicted, however, VO2max was > or = 69% in all 5 patients, indicating only mild impairment of exercise capacity in the 2 patients with < 15 ml/kg/min VO2max. Therefore all 5 patients were offered surgery and underwent lobectomy. Apart from 1 prolonged air leak no complications occurred, the mean hospital stay was 16 days (13-21).(ABSTRACT TRUNCATED AT 250 WORDS)