AIM:To investigate the contribution of fluorescein angiographic leaking microaneurysms ( leak-MA) versus non- leaking microaneurysms ( non-leak-MA) to retinal thickening in diabetic retinopathy. METHODS:A consecutive series of 38 eyes from 24 patients with diabetic retinopathy was included. Leak-MA and non-leak-MA in each eye were selected in pairs at corresponding topographic location. Leaking was defined by late phase fluorescein angiograms compared to early phase. Retinal thickness was measured with Heidelberg Spectralis OCT topographically aligned on early phase angiograms at the MA site and within a 1 mm circle. RESULTS:In all eyes,significant retinal thickening at the site of leaking compared to non-leaking microaneurysms was observed ( 356 ±69μm vs 318 ±56μm,P <0.001) ,showing a mean increase in thickness in the areas of leak-MA vs non- leak-MA of 38 ± 39μm ( 95% confidence interval 25-51μm,P < 0. 001 ) . All 1mm area measurements also showed significant ( P <0.001) thickening of the leak-MA with average thickness of leak-MA vs non-leak-MA as 351 ± 67μm vs 319 ± 59μm; minimum thickness 311 ±62μm vs 284 ±60μm; maximum thickness 389 ±78μm vs 352 ±66μm; and retina volume 26.4 ±6. 0mm3 vs 23.6 ±3.7mm3,respectively. CONCLUSION:Leaking of microaneurysms on fluorescein angiography appears to cause focal thickening of retina,which can be measured with high-resolution OCT. Therefore, targeting leaking microaneursyms in diabetic retinopathy has the potential to reduce retinal thickening.