Background: The accuracy and convenience of pen devices for insulin injection have improved quality of life for patients with insulin-treated diabetes mellitus (DM). Prefilled, disposable pens have the advantage of simplicity, with minimal training and attention required and no installation of new cartridges necessary. Objective: The aim of this study was to assess patient preference, efficacy, and safety profiles of a prefilled, disposable pen (FlexPen®) and conventional vial/syringe injection method for insulin injection therapy among patients with DM. Methods: In a multicenter, randomized, open-label, crossover study, patients with type 1 or 2 DM were transferred from previous QD or BID conventional insulin therapy to a mixture of 70% insulin aspart protamine suspension and 30% insulin aspart injection (NovoLog® Mix 70 30 ) for 4 weeks of dose optimization using their usual type of syringe. Patients were then randomly assigned to use either vial/syringe or a prefilled, disposable pen to inject the biphasic insulin aspart 70 30 mixture for the next 4 weeks, followed by 4 weeks of use of the other injection device. Efficacy, safety profiles, and patient preference for the delivery systems were compared. Results: A total of 121 patients (mean [SD] age, 57.0 [12.4] years; age range, 28–81 years; mean [SD] body mass index, 31 [5.5] kg/m 2) were enrolled. One hundred three patients completed the study. Seventy-four percent of patients ( 78 105 ) indicated a preference for the pen over the vial/syringe method (95% CI, 71%–87%), compared with 20% ( 21 105 ) who preferred the vial/syringe. Eighty-five percent ( 88 104 ) considered the pen more discreet for use in public (compared with 9% [ 9 104 ] for the vial/syringe), 74% ( 77 104 ) considered it easier to use overall (compared with 21% [ 22 104 ] for the vial/syringe), and 85% ( 89 105 ) found the insulin dose scale on the pen easier to read (compared with 10% [ 10 105 ] for the vial/syringe). Patients had statistically significant improvement in glycosylated hemoglobin values during the study ( P < 0.05). No statistically significant differences in fasting plasma glucose, mean 4-point blood glucose profiles, or serum fructosamine values were found between groups. Overall, the safety profiles during treatment periods with the pen were comparable to those with the vial/syringe. Conclusions: In this trial, differences in efficacy and safety profiles between the vial/syringe and prefilled, disposable pen appeared negligible. However, more patients expressed a preference to continue use of the pen.