BACKGROUND: Insulin initiation and optimization is a challenge for patients with type 2 diabetes. Our objective was to determine whether safety and efficacy of AIR inhaled insulin (Eli Lilly and Co., Indianapolis, IN) (AIR is a registered trademark of Alkermes, Inc., Cambridge, MA) using a simplified regimen was noninferior to an intensive regimen. METHODS: This was an open-label, randomized study in insulin-naive adults not optimally controlled by oral antihyperglycemic medications. Simplified titration included a 6 U per meal AIR insulin starting dose. Individual doses were adjusted at mealtime in 2-U increments from the previous day's four-point self-monitored blood glucose (SMBG) (total < or =6 U). Starting Air insulin doses for intensive titration were based on fasting blood glucose, gender, height, and weight. Patients conducted four-point SMBG daily for the study duration. Insulin doses were titrated based on the previous 3 days' mean SMBG (total < or =8 U). RESULTS: End point hemoglobin A1C (A1C) was 7.07 +/- 0.09% and 6.87 +/- 0.09% for simplified (n = 178) and intensive (n = 180) algorithms, respectively. Noninferiority between algorithms was not established. The fasting blood glucose (least squares mean +/- standard error) values for the simplified (137.27 +/- 3.42 mg/dL) and intensive (133.13 +/- 3.42 mg/dL) algorithms were comparable. Safety profiles were comparable. The hypoglycemic rate at 4, 8, 12, and 24 weeks was higher in patients receiving intensive titration (all P < .0001). The nocturnal hypoglycemic rate for patients receiving intensive titration was higher than for those receiving simplified titration at 8 (P < 0.015) and 12 weeks (P < 0.001). CONCLUSIONS: Noninferiority between the algorithms, as measured by A1C, was not demonstrated. This finding re-emphasizes the difficulty of identifying optimal, simplified insulin regimens for patients.