This paper describes an approach that successfully classifies five types of white blood cells based on image processing algorithms and a Support Vector Machine classifier. The images were taken from peripheral blood of healthy subjects. The cells, nucleus, and cytoplasm are segmented through thresholding techniques and mathematical morphology. Then, a set of textural and morphological features are extracted from the segmented regions. System evaluation is performed using 4-fold stratified cross-validation. An overall accuracy of 90.5 ± 1.6% was achieved.