PURPOSE: To prospectively compare the efficacy and safety of 4 vs 2 mg intravitreal triamcinolone acetonide (TA) injection for diabetic macular edema. DESIGN: Interventional case series. METHODS: PATIENTS: Thirty-two patients with diabetic macular edema unresponsive to laser photocoagulation. INTERVENTION: Patients were randomly assigned to re, ceive 4 or 2 mg intravitreal TA in one eye (16 patients in each group). MAIN OUTCOME MEASURES: The main outcome was central macular thickness (CMT) measured by optical coherence tomography (OCT) at four, 12, and 24 weeks. Secondary outcomes were gain in Early Treatment Diabetic Retinopathy Study (ETDRS) scores, intraocular pressure (IOP), cataract progression, and duration of effect. RESULTS: Before injection, mean (+/- SD) CMT was 564.5 +/- 119 mu m and 522.9 +/- 148.5 mu m in the 4- and 2-mg groups, respectively. At four, 12, and 24 weeks after injection, it was 275.0 +/- 79.8, 271.4 +/- 128.7, and 448.7 +/- 146.4 mu m, respectively, in the 4-mg group, and 267.3 +/- 82.4, 289.8 +/- 111.4, and 394.7 +/- 178.9 mu m, respectively, in the 2-mg group. At no time was the difference in CMT between both groups statistically significant (P > 0.3). The between-group differences in the gain in the ETDRS score and in IOP were not statistically significant either. Diabetic macular edema recurred after a median period of 20 weeks vs 16 weeks in the 4- and 2-mg groups, respectively (P = 0.11). CONCLUSIONS: In the short term, intravitreal injection of 4 or 2 mg TA does not have different effects on CMT, visual acuity, or IOP.