Transcription factors of the GATA-family are essential for proper development of diverse tissues or cell types. GATA-1 is required for differentiation of two hematopoietic lineages (red blood cells and megakaryocytes), whereas GATA-3 is essential for T-cell development. Functional studies suggest that many properties of the GATA-family of proteins are shared and largely interchangeable. To test whether the function of GATA-1 in erythroid differentiation can be replaced by another GATA-factor, we generated a knock-in mutation of the GATA-1 locus in which GATA-3 cDNA was introduced by gene targeting. Mutant embryos (designated G1G3ki), though embryonic lethal, exhibit partial rescue, characterized by increased survival of erythroid precursor cells and improved hemoglobin production. The basis for the incomplete extent of rescue is likely to be complex, but may be accounted for, in part, by insufficient accumulation of GATA-3 protein (compared with the normal level of GATA-1). Our findings suggest that GATA-3 protein is functional when expressed in an erythroid environment and is competent to act on at least a subset of erythroid-expressed target genes in vivo.