BACKGROUND The purpose of the study is to determine whether the pre-treatment clinical systemic variables and optical coherence tomography (OCT) findings are associated with the subsequent response to the intravitreal bevacizumab (IVB) in eyes with persistent diabetic macular edema (DME). DESIGN Prospective, interventional non-comparative case series study. METHODS 38 Patients (45 eyes) with refractory diabetic macular edema; 16 females, 22 males with a mean aged 57.5 year. All patients had persistent DME not responded to other forms of treatments. Complete eye examination; best corrected visual acuity (BCVA) (represented as LOGMAR for adequate statistical analysis), slit-lamp exam, intraocular pressure measurement, stereoscopic biomicroscopy of the macula, and morphologic patterns of diabetic macular edema demonstrated by OCT. All patients had intravitreal injection of 0.05 mL = 1.25 mg bevacizumab (Avastin; Genentech, Inc., San Francisco, CA), and were followed up for 3 months. The pre and post-injection follow-up data were analyzed by Student-t test and Mann-Whitney test for two main outcome measures; visual acuity (LOGMAR) and central foveal thickness (CFT) changes over a period of three months, and the data included demographic factors, type, duration and the control of diabetes mellitus (HbA1C%), grade of diabetic retinopathy, renal function (serum creatinine level), serum cholesterol, blood pressure control and previous treatment by focal laser and/or intravitreal triamcinolone injection. RESULTS The LOGMAR and central foveal thickness (CRT) improved in 30/45 eyes (67%) and 32/45 eyes (72%), respectively during a mean follow-up time of three months. The mean LOGMAR visual acuities were 0.64 (SD ± 0.34), 0.61 (SD ± 0.31) and 0.60 (SD ± 0.32) at pre-injection, at 1 month post-injection and at 3 months post-injection, respectively; but this mean increase in vision was statistically not significant (P value = 0.099). The mean foveal thicknesses were 444.95 μm (SD ± 127.36), 394.95 μm (SD ± 138.03) and 378.32 μm (SD ± 112.01) at pre-injection, 1 month post-injection and 3 months post-injection, respectively, this decrease in the foveal thickness was statistically significant (P value <0.001). The pre and post-injections values of the variables for diabetic duration, diabetic control (HbA1c) and OCT pattern of macular edema showed significant statistical correlations (P < 0.05) with LOGMAR only, however the values of the variables for serum creatinine and cholesterol show statistical correlation (P < 0.05) with both LOGMAR and CFT. CONCLUSIONS Chronicity and inadequate control of diabetes mellitus, nephropathy, hyperlipidemia and presence of vitreomacular attachment (VMA) are factors associated with poor vision progress after intravitreal bevacizumab injection.