AIM: To compare the effectiveness of intravitreal bevacizumab and subthreshold macular photocoagulation (SMP), for the treatment of non-center involved diabetic macular edema (non-CI DME). METHODS: Prospective, randomized, controlled clinical study included patients with type 2 diabetes, non-CI DME and best-corrected visual acuity (BCVA) of 0.30 logMAR or better. Each eye was randomized into three groups: group 1, monthly intravitreal bevacizumab; group 2, single SMP; group 3, single SMP and monthly bevacizumab. Main outcome measures were BCVA, and macular thickness measured with optical coherence tomography as macular central subfield thickness (CST), macular area of greater thickness (MAGT) and total macular volume (TMV). Results were analyzed after 3mo. RESULTS: A total of 32 eyes were included. Group 3 improved in BCVA (0.19±0.16 to 0.12±0.14 logMAR; P=0.041) and in TMV (7.90±0.57 to 7.65±0.73 mm3; P=0.025). Group 1 improved in MAGT (325±26.26 to 298.20±44.85 μm; P=0.022) and TMV (7.79±0.57 to 7.50±0.56 mm3, P=0.047). Group 2 didn’t show significant improvement of any variable. CONCLUSION: The loading phase of bevacizumab as monotherapy or combined with SMP is superior to SMP as monotherapy in providing short-term visual and anatomical improvement in non-CI DME.