Image filtering is an important technology used to eliminate the noise of distorted images in image processing applications. Digital images are often distorted by noise during shooting, and so their original features may not be observed. For this reason, noise reduction is an important step for the analysis of noisy images produced by digital camera flaws and during transmission of data. In this study, fuzzy set and 3×3, 3×5, 2×3 size filtering windows and mean - median filters were used to reduce salt-pepper noise. The performances of the proposed methods are then compared with some methods in the literature according to Peak Signal-to-Noise Ratio (PSNR) and Mean Square Error (MSE).