OBJECTIVE: Glycemic control using inhaled, dry-powder insulin plus a single injection of long-acting insulin was compared with a conventional regimen in patients with type 2 diabetes, which was previously managed with at least two daily insulin injections. RESEARCH DESIGN AND METHODS: Patients were randomized to 6 months' treatment with either premeal inhaled insulin plus a bedtime dose of Ultralente (n = 149) or at least two daily injections of subcutaneous insulin (mixed regular/NPH insulin; n = 150). The primary efficacy end point was the change in HbA1c from baseline to the end of study. RESULTS: HbA1c decreased similarly in the inhaled (-0.7%) and subcutaneous (-0.6%) insulin groups (adjusted treatment group difference: -0.07%, 95% CI -0.32 to 0.17). HbA1c < 7.0% was achieved in more patients receiving inhaled (46.9%) than subcutaneous (31.7%) insulin (odds ratio 2.27, 95% CI 1.24-4.14). Overall hypoglycemia (events per subject-month) was slightly lower in the inhaled (1.4 events) than in the subcutaneous (1.6 events) insulin group (risk ratio 0.89, 95% CI 0.82-0.97), with no difference in severe events. Other adverse events, with the exception of increased cough in the inhaled insulin group, were similar. No difference in pulmonary function testing was seen. Further studies are underway to assess tolerability in the longer term. Insulin antibody binding increased more in the inhaled insulin group. Treatment satisfaction was greater in the inhaled insulin group. CONCLUSIONS: Inhaled insulin appears to be effective, well tolerated, and well accepted in patients with type 2 diabetes and provides glycemic control comparable to a conventional subcutaneous regimen. Copyright 2004 American Diabetes Association