The role of non-invasive nocturnal domiciliary ventilation (NNV) in chronic obstructive pulmonary disease (COPD) patients with chronic hypercapnia is still discussed. The aims of this study were to evaluate the long-term survival, the clinical effectiveness and side-effects of NNV in these patients. Forty-nine stable hypercapnic COPD patients on long-term oxygen therapy (LTOT) were assigned to two groups: in Group 1, 28 patients performed NNV by pressure support modality in addition to LTOT; in Group 2, 21 patients continued their usual LTOT regimen. Treatment was assigned according to the compliance to NNV, after an in hospital period. Mortality rate, hospital stay (HS) and ICU admissions (IA) were recorded in the two groups. HS and IA were compared to those recorded in a similar period of follow-back. Lung and respiratory muscle function, dyspnoea, and exercise capacity (by 6-min walk test) were evaluated baseline and every 3-6 months up to 3 yr. Mean follow-up time was 35 +/- 7 months. Mortality rate was not different between the two groups: 16, 33, 46% and 13, 28, 50% at 1, 2 and 3 yr in Groups 1 and 2 respectively. Lung and respiratory muscle function did not significantly change over time. A significant increase in 6-min walk test (from 245 +/- 78 to 250 +/- 88, 291 +/- 75, 284 +/- 89 m after 1, 2 and 3 yr respectively, P < 0.01) was observed only in patients undergoing NNV. In comparison to the follow back HS significantly decreased in both groups (from 37 +/- 29 to 15 +/- 12 and from 32 +/- 18 to 17 +/- 11 days/pt/yr in Groups 1 and 2 respectively, P < 0.001) whereas IA significantly decreased only in patients performing also NNV (from 1.0 +/- 0.7 to 0.2 +/- 0.3/pt/yr, P < 0.0001). Addition of NNV by pressure support modality to LTOT does not improve long term survival but significantly reduces ICU admissions and improves exercise capacity in severe COPD with hypercapnia.