BACKGROUND: Several lines of preclinical studies have shown promising antidiabetic effects of the aqueous leaves extract of Coccinia grandis (Linn.) Voigt (Cucurbitaceae) in vivo and in vitro. PURPOSE: The present study was conducted to evaluate the efficacy and safety of a newly developed herbal formulation of C. grandis in newly diagnosed patients with type 2 diabetes mellitus (T2DM). STUDY DESIGN: A three months long, randomized, double blind, placebo controlled clinical trial in patients with newly diagnosed T2DM. METHOD: Based on fasting plasma glucose (FPG) concentration, a total number of 158 newly diagnosed patients with T2DM (45 ± 15 years age) were recruited for the present trial from the University Medical Clinic, Teaching Hospital, Karapitiya, Galle, Sri Lanka. They were randomly assigned to the test or placebo group to receive 500 mg of herbal drug (n = 79) or placebo drug (n = 79) once daily for three months. Patients and investigators were blinded for the treatment. Percentage of glycated hemoglobin (HbA %), insulin and lipid profile parameters were estimated at the base line and at the end of the intervention. Serum concentration of fructosamine was assessed at every other visit of the trial. The homeostatic model assessment for insulin resistance (HOMA-IR), atherogenic index (AI), cardio-protective index (CPI) and coronary risk index (CRI) were calculated. Furthermore, fasting plasma glucose concentration, renal and liver toxicity parameters, hematological parameters, blood pressure (BP) were assessed throughout the study in two weekly intervals till the end of three months. RESULTS: Out of 158, a total number of 145 patients completed the entire clinical trial period successfully. Mean (SD) changes of variables from the baseline to the end of the intervention in test and placebo groups were 0.65 (0.54) and 0.08 (0.66) for HbA % (p < 0.001), 1.91 (3.07) and -1.28 (9.77) for insulin (p < 0.001), 0.02 (0.03) and -0.01 (0.04) for frucosamine (p < 0.001), 1.51 (0.49) and 0.05 (0.50) for FPG (p < 0.001), 1.73 (1.36) and -0.37 (3.38) for HOMA-IR (p < 0.001), 0.16 (0.18) and -0.04 (0.42) for TG (p < 0.001), 0.07 (0.08) and -0.02 (0.19) for VLDL-C (p < 0.001), respectively. However, the herbal drug of C. grandis was unable to change other outcome variables significantly when compared to the placebo (p > 0.05). All the renal, liver and toxicity parameters, hematological parameters and BP were within the normal physiological reference ranges at each visit. CONCLUSION: Treatment with herbal drug of C. grandis (500 mg per day) for three months for patients with newly diagnosed T2DM significantly improved their glycemic and selected lipid profile parameters with well tolerated safety.