We have performed a retrospective survey on 296 patients, who attended the Respiratory Function Unit at Guy's Hospital to have their bronchodilator responses (BR) tested. The aim of the study was to see the effect of ipratropium bromide (IB) in a group of patients with incomplete reversibility after salbutamol (S). Patients were routinely given salbutamol and ipratropium bromide sequentially by inhalation, and spirometric changes were recorded after each drug. We identified two groups: Group A, 95 patients with FEV1 response greater than or equal to 0.2 l after either drug and FEV1 less than 80% predicted after salbutamol; and, Group B, 49 with change in FEV1 less than 0.02 l, FVC less than 80% predicted after salbutamol and an improvement in FVC greater than or equal to 0.33 l. Seventy-nine of the 95 patients in Group A also had an FVC response. In Group A, age was negatively correlated with response to salbutamol (r = -0.41, P less than 0.0001), and within Group B baseline FVC was negatively correlated with response to ipratropium bromide (r = -0.30, P = 0.03). There were no differences in age, sex, or doses given to each group (median dose: salbutamol, 800 micrograms, ipratropium bromide 120 micrograms). Baseline FEV1 and FVC (% predicted) were significantly higher in FEV1 responders. Mean (SD) FEV1 were 43% (14) in Group A vs. 29% (14) in Group B, while FVC were 62% (16) vs. 47% (13), P less than 0.001. Responses to ipratropium bromide were more frequent in Group B; in Group A 87% improved after salbutamol and 26% after ipratropium bromide, while in Group B 68% responded to salbutamol and 47% to ipratropium bromide (P = 0.03). Most patients responded to salbutamol, but in 33% ipratropium bromide had an additional effect. The FEV1 response to salbutamol declined with age. Isolated volume responders had more severe airflow obstruction, had less responses to salbutamol and were more likely to show a response to ipratropium bromide. These results support a trial of ipratropium bromide in patients with inadequate beta responsiveness, especially in those with severe airflow obstruction.