PURPOSE: To determine precisely the mean change in refractive power induced by treatment in patients with diabetic macular edema (DME). DESIGN: Prospective, randomized study. PARTICIPANTS: Fifty eyes of 50 consecutive patients with clinically significant macular edema receiving all 3 types of current state-of-the-art treatment with intravitreal antiedematous substances (ranibizumab, bevacizumab, or triamcinolone). METHODS: Patients were followed up at monthly intervals and were treated following a standardized pro re nata regimen according to protocol. Best-corrected visual acuity (BCVA) was determined by certified visual acuity examiners. The refractive power of the treated eyes was determined using a push-plus technique. The change in refraction between baseline and the visit when the macula was completely dry or when the central subfield thickness (CST) measured by optical coherence tomography had reached the thinnest level was analyzed. MAIN OUTCOME MEASURES: Spherical equivalent refraction (SER) and CST. RESULTS: Fifty eyes of 50 patients received intravitreal therapy using ranibizumab (n = 11), bevacizumab (n = 20), or triamcinolone (n = 19). Mean BCVA was 0.33±0.23 logarithm of the minimum angle of resolution (logMAR) and mean CST was 492±130 μm. The mean SER was 0.41±2.06 diopters (D) at baseline. The BCVA at the time of optimal retinal morphologic features was 0.24±0.2 logMAR, mean CST was 300±78 μm, and mean change in SER was -0.01±0.46 D. Changes is BCVA and CST were statistically significant (P < 0.0001), but the SER change was not (P = 0.824). CONCLUSIONS: Appropriate spectacle correction can be prescribed to patients with DME any time during ongoing therapy using antiedematous substances because resolution of retinal thickening is not associated with an increased risk of a myopic shift.