Decreased chemosensitivity to hypercapnia, a common finding in Type 1 diabetes mellitus, seems related to autonomic neuropathy. We proposed to verify whether simple neuroautonomic cardiovascular tests or indexes of severity of diabetes and respiratory impairment can identify patients with such a dysfunction, but no clinical evidence of autonomic neuropathy. Forty patients with Type 1 diabetes, 20 with autonomic neuropathy according to the results of a standardized test battery, were studied and compared with 40 normal subjects matched by age and sex. Spirometry and pulmonary diffusing capacity for carbon monoxide were performed. The chemosensitivity to hypercapnia was tested by the rebreathing method. There was no significant difference between patients with and without autonomic neuropathy in chemosensitivity to hypercapnia, as expressed by the ventilation response to increasing end-tidal pressure of carbon dioxide; however, it was lower in the whole group of patients with diabetes than in control subjects (1.71 ± 0.80 vs. 2.45 ± 1.11 l⁻¹ min⁻¹ mmHg, respectively, P=0.002). No significant correlation was found between ventilation response to increasing end-tidal pressure of carbon dioxide and the results of autonomic tests. In patients with diabetes mellitus, the ventilatory response to hypercapnia significantly correlated with pulmonary diffusing capacity for carbon monoxide (Spearman's rho=0.387, P=0.013) and this was the only variable significantly associated with ventilation response to increasing end-tidal pressure of carbon dioxide in a multiple regression model. Chemosensitivity to hypercapnia was depressed in patients with diabetes mellitus, irrespective of autonomic neuropathy, in comparison with control subjects. The correlation with pulmonary diffusing capacity for carbon monoxide suggests that microcirculatory damage might contribute to depress the central chemosensitivity.