OBJECTIVE To investigate whether changing the prandial regular insulin to rapid-acting insulin analogue in hospital medicine wards improves the timing of insulin delivery in relation to meals and improves patient safety and glucose control. METHODS This open-label randomized controlled trial in type 2 diabetic patients compared insulin lispro with meals and basal insulin glargine (intervention) vs regular insulin before meals and basal neutral protamine Hagedorn insulin twice daily (control). The primary endpoint was the rate of targeted timing of insulin to meals (target time). In the intervention group, target time was defined as insulin administered from 15 minutes before to 15 minutes after the patient started a meal. For the control group, target time was defined as insulin administered from 30 minutes before to 30 minutes after the patient started a meal. Hypoglycemic, hyperglycemic, and severe hyperglycemic patient-days were compared between groups. RESULTS Twenty-seven patients in the intervention group and thirty-three patients in the control group were studied. The percentage of times that the insulin was given within target time was significantly higher in the intervention group as a whole (88.9% vs 70.1%, P<.001) and was higher for lunch and the evening meal (90% vs 66.7% and 94.7% vs 70.1%, P<.001). The rate of hypoglycemia was lower in the intervention group (1.85% vs 15%, P<.001). The rate of hyperglycemia was similar in both groups (68.2% vs 59.8%, P = .224), but the intervention group had a higher rate of severe hyperglycemia (28.9% vs 12.9%, P = .003). CONCLUSIONS The use of prandial insulin analogues in medicine wards allows better timing with meals than regular insulin and results in better hypoglycemic outcomes. Higher rates of hyperglycemia with prandial analogues may need adjustment in insulin doses.