Cell and particle quantification is one of the frequently used techniques in biology and clinical study. Variations of cell/particle population and/or protein expression level can provide information on many biological processes. In this chapter, we propose an image-based automatic quantification approach that can be applied to images from both fluorescence and electron microscopy. The algorithm uses local maxima to identify labelling targets and uses watershed segmentation to define their boundaries. The method is able to provide information on size, intensity centroids and average intensity within the labelling partitions. Further developed from this method, we demonstrated its applications in four different research projects, including recruitment enumeration of circulating T cell in non-lymphoid tissues, cell clustering in the early development of the chick embryo, gold particle localization and clustering in electron microscopy, and registration/co-localization of transcription factors in neural tube development of early chick embryo. The advantages and limitations of the method are also discussed.