Objective. To evaluate the efficacy of Abelmoschus manihot in treating type 2 diabetic nonproliferative retinopathy. Methods. It was a randomized controlled clinical trial. The recruited eighty subjects with type 2 diabetic nonproliferative retinopathy were randomly divided into treatment group and control group. The two groups received basic treatments including control of blood glucose, blood pressure and blood lipid, management of diet, exercise and health education, and monitoring of relevant indicators. Additionally, the treatment group was given oral administration of Abelmoschus manihot. All subjects were followed up on monthly basis for consecutive six months. The related parameters including diabetic retinopathy (DR) incidence rates, “Early Treatment Diabetic Retinopathy Study” (ETDRS) vision scores, retinal thicknesses in macular region, serum vascular endothelial growth factor (VEGF) levels, and biochemical indicators of both groups before and after treatment were accurately collected and statistically analyzed. Results. There were no significant differences of DR severity levels, ETDRS vision scores, macular retinal thicknesses such as cube average thickness (CAT), central subfield thickness (CST), and cube volume (CV), and serum VEGF levels between two groups before treatment. Meanwhile, there were no significant differences of demographic characteristics, case terminations, blood glucose, blood lipid, blood pressure, biochemical indicators of hepatorenal function, hypoglycemic drugs, hypotensive drugs, and other basic treatments between two groups during six months treatment. The present study suggested that the remission rate of DR and the ETDRS vision score in the treatment group were significantly higher than those of the control group (remission rate: 25.4% vs 9.3%, P=0.01; ETDRS score: 78 (72, 82) vs 72 (67, 80), P=0.0002) while the progression rate of DR in the treatment group was significantly lower than that of the control group (progression rate: 4.2% vs 18.7%, P=0.007) after six months treatment. In addition, the CAT, CST, CV, and serum VEGF levels of the treatment group were significantly improved after the treatment (CAT: 286 (278, 302) vs 282 (270, 295) μm, P<0.0001; CST: 251 (239, 274) vs 248 (235, 265) μm, P<0.0001; CV: 10.3 (10.0, 10.9) vs 10.1 (9.7, 10.6) mm3, P<0.0001; VEGF: 0.21 (0.14, 0.58) vs 0.16 (0.10, 0.23) ng/ml, P=0.0026), while there were no significant differences of the control group before and after treatment (CAT: 287 (279, 294) vs 287 (279, 295) μm, P=0.27; CST: 250 (240, 266) vs 252 (238, 266) μm, P=0.72; CV: 10.4 (10.1, 10.6) vs 10.4 (10.1, 10.7) mm3, P=0.53; VEGF: 0.21 (0.13, 0.66) vs 0.23 (0.12, 0.64) ng/ml, P=0.85). Conclusion. The study offered the novel evidence for the therapeutic effect of Abelmoschus manihot on type 2 diabetic nonproliferative retinopathy, which was associated with improved VEGF. This trial is registered with ChiCTR1800019292.