Aim: To compare insulin and GLP-1 analogues therapy on glycemic control in poorly controlled T2DM subjects failing on oral therapy. Methods: The electronic database PubMed was systematically searched for randomized controlled trial (RCTs) with duration>16 weeks comparing the addition of insulin therapy versus GLP-1 analogues in poorly controlled T2DM subjects on oral therapy. Results: We identified 7 RCTs with 2199 patients of whom 1119 were assigned to insulin therapy and 1080 received a GLP-1 analogue. Both insulin and GLP-1 analogues were effective in lowering HbA1c with no statistically significant difference between the mean decreases in HbA1c. However, insulin was more effective than GLP-1 analogues in lowering the fasting plasma glucose (FPG) concentration, while GLP-1 agonists were more effective in lowering the post prandial glucose (PPG) concentration. Insulin therapy was associated with weight gain while GLP-1 analogues consistently caused weight loss and the difference between the mean change in body weight between the two therapies was highly statistically significant. Despite a similar decrease in HbA1c, the risk of hypoglycemia was 35% lower (p=0.001) with GLP-1 therapy compared to insulin. Compared to insulin, GLP-1 analogues caused a significant decrease in systolic blood pressure and were associated with greater rate of gastrointestinal adverse events. Conclusion/interpretation: In poorly controlled T2DM subjects on oral therapy, GLP-1 analogues and insulin are equally effective in lowering the HbA1c. However, GLP-1 analogues have additional nonglycemic benefits and lower risk of hypoglycemia. Thus, GLP-1 analogues should be considered as a treatment option in this group of diabetic individuals.