To build a coherent picture of mitosis and cell fates during blastoderm and through the complex movements of gastrulation, it will be important to localize and follow several markers simultaneously in live specimens, ideally in 3D, using high-resolution, specific, noninjurious staining and observation procedures. The study of early Drosophila development has already profited from the use of fluorescent labeling and low-light-level imaging of live embryos using a CCD camera. Chromosomes in fixed samples have been labeled using DNA-specific dyes, making the pattern of mitotic patches visible. In vivo, 3D microscopy of fluorescently tagged chromosomes, in conjunction with computerized image processing, has permitted the first direct cell lineage analysis in the early Drosophila embryo. Moreover, the techniques adapted to study Drosophila development have been used for analysis of Drosophila chromosome structure, mitosis, and cell cycle, and are general enough to be applied to a myriad of problems in cell biology. "Optical sectioning" has always been used to scrutinize everything from onion roots to frog eggs, focusing up and down through the specimen, with the observer's brain responsible for the image processing. However, the volume of raw data generated by the high-resolution approach detailed above requires the use of sophisticated and adaptable computer systems to analyze and organize the results. Software designed to extract information from these complex images, either automatically or through an interactive approach, will become essential tools for cell and developmental biology. The brain of the experimenter remains the most important component in any image-processing system, but the support of technology will be essential.