The ancestry of dopaminergic (DA) neurons in the Xenopus laevis hypothalamus was investigated by combining intracellular lineage dye injections of 16- and 32-cell blastomeres with the immunofluorescent detection of tyrosine hydroxylase at tadpole stages. At these stages, DA neurons in the hypothalamus comprise a discrete nucleus that contains from 22 to 45 cells on each side [mean = 32.6 +/- 6.6 (SD)]. The DA nucleus descends from only four of the 16-cell blastomeres. The two dorsal midline blastomeres (D1.1) are the major progenitors, and in all embryos studied they contributed to the DA nucleus. The two dorsal lateral blastomeres (D1.2) contribute to the DA nucleus in only about half of the embryos. Thus, the DA nucleus descends only from a discrete group of progenitors, and the participation of some of the progenitors in the DA lineage is only probabilistic. The number of DA neurons generated by the same blastomere varied greatly in different animals. This variation in cell number correlated with the degree of coherence and the density of the clone in the hypothalamus, rather than with clonal ancestry. Bilateral deletion of the major 32-cell progenitor (D1.1.1) resulted in a nearly complete restitution of the DA nucleus in 74% of the embryos that successfully completed gastrulation and neurulation. In the rest, the hypothalamus was smaller than normal or missing, and the DA nucleus was significantly reduced in size or absent. These results show that the DA nucleus can be restored after its normal lineage is deleted, but complete regulation is not always accomplished. Several blastomere progenitors dramatically altered their contribution to the DA nucleus after D1.1.1 ablation, including two blastomeres that normally do not contribute to the DA lineage. Thus, the fate to produce DA neurons is not determined at cleavage stages.