Exudates are one of the primary signs of diabetic retinopathy, which is a main cause of blindness and can be prevented with an early screening process. In this paper, authors have attempted to detect exudates using back propagation neural network. The publicly available diabetic retinopathy dataset DIARETDB1 has been used in the evaluation process. To prevent the optic disk from interfering with exudates detection, the optic disk is eliminated. Significant features are identified from the images after preprocessing by using two methods: Decision tree and GA-CFS method are used as input to the BPN model to detect the exudates and non-exudates at pixel level. The results prove that, BPN performance with features identified by Decision tree and GA_CFS approach has outperformed the performance of BPN with all inputs. The BPN classifier best performance was found with Sensitivity of 96.97 %, Specificity of 100% and classification accuracy of 98.45%.