Erythropoiesis, measured by 24-hr Fe59 uptake by spleens and erythrocytes, was studied in mice of several genotypes 1 -2 weeks after irradiation and transplantation of bone marrow cells from isogenic or allogenxc mice. When 2 parental strains, B6 and D2, and their Fj hybrid offspring were compared in the 9 possible donorhost combinations, erythropoiesis of the 3 groups (F1[forward arrow] F1, B6[forward arrow] B6, and D2 [forward arrow] F1) was significantly greater, and of three others (B6[forward arrow] D2, B6 [forward arrow] F1, and D2 [forward arrow] B6) significantly less, than that of the remaining groups. B6 and D2 marrow tested separately in recipients of several other genotypes gave rise to variable erythropoiesis, the extent of which could not be simply correlated with H-2 homozygosity or type of the recipient. Marrow from male and female donors produced erythrocytes equally well in recipients of either sex. The data show that growth of cellular transplants in an irradiated host cannot be predicted accurately from knowledge of H-2 disparities alone. They do not offer any support for an immunological interpretation of poor growth of parental cells in F1 hybrids.