The fates of lineage labeled hematopoietic precursor populations in Xenopus embryos are followed by use of in situ hybridization, looking for overlap between lineage labeled cells and in situ probes specific for known cell populations or states of differentiation. By coinjection of dominant interfering constructs, it also is possible to define the environmental cues or signals required for specification and/or maintenance of the hematopoietic program at different times and locations in the early embryo. As a lineage trace, we use beta-galactosidase, which is injected as in vitro synthesized ribonucleic acid (RNA) in to Xenopus embryos at early cleavage stages. Because the interfering constructs we use also are in the form of injected RNA, the use of beta-galactosidase RNA as a lineage trace assures accurate determination of the cells expressing the dominant negative construct. Embryos are cultured to desired developmental stages, fixed briefly and processed for the beta-galactosidase reaction. Embryos are then analyzed by whole mount in situ hybridization, embedded in wax, and sectioned. Alternatively, after the beta-galactosidase reaction, embryos can be fixed long term in paraformaldehyde, mounted in wax, sectioned, and probed by in situ hybridization directly on sections.