Lipid profile represents a driver of cardiovascular risk in type 2 diabetes. The effect of chronic insulin therapy on cholesterol levels is unclear. We aim to evaluate the effect of basal insulin on lipid profile compared to other classes of anti-hyperglycaemic agents in type 2 diabetic patients.We performed a meta-analysis of randomized controlled trials reporting changes of lipid parameters in type 2 diabetic patients randomized to basal insulin or other classes of anti-hyperglycaemic agents.The levels of total (TC) and LDL cholesterol (LDL-C) appeared to be significantly reduced by therapies with glucagon-like peptide-1 receptor agonists (GLP-1RA) in comparison to basal insulin (mean difference [MD] -3.80; 95% CI [-6.30 to -1.30] mg/dl, p<0.001 and -4.17; 95% CI [-6.04 to -2.30] mg/dl, p<0.0001), whereas no difference was detected between basal insulin and dipeptidyl peptidase-4 inhibitors (DPP4-I) or standard therapy (sulfonylurea ± metformin). Thiazolidinediones (TZD) produced a significant improvement in HDL cholesterol (HDL-C) (MD 3.55; 95% CI [0.55 to 6.56] mg/dl, p=0.02) but were associated with an increase in TC and LDL-C (MD 16.20; 95% CI [9.09 to 23.31] mg/dl, p<0.001 and 5.19: 95% CI [-3.00 to 13.39] mg/dl, p=0.21). Basal insulin was superior to standard therapy in triglyceride reduction (MD 3.8; 95% CI [0.99 to 6.63] mg/dl, p=0.008).GLP-1RA were superior to basal insulin in the control of TC and LDL-C. Basal insulin effectively reduced serum triglycerides. TZD led to improvement in HDL-C. DPP4-I and standard therapy did not have any significant effect on lipid levels.