Butyric acid, a naturally occurring fatty acid, has been shown to increase fetal hemoglobin in BFUe cultures, in primates, and in patients with beta chain hemoglobinopathies. The precise mechanism of gamma gene induction by butyrate is unknown. Butyrate may induce fetal hemoglobin production in vivo by reactivation of silenced gamma globin genes, by inhibiting the silencing of gamma genes, or by both mechanisms. We examined the effects of butyrate on gamma gene expression in transgenic mice carrying three types of constructs: microLCRA gamma mice, which continue to express the gamma gene in the adult stage of development at a level of one-third to one-fifth of the expression in the fetus; microLCRA gamma psi beta delta beta mice, which display correct developmental regulation of gamma and beta human globin genes and have low level gamma globin expression in the adult; and beta locus YAC mice, which display correct developmental regulation of epsilon, gamma, and beta globin genes and have a totally silenced gamma gene in the adult stage. Animals were treated with a continuous infusion of alpha-amino butyric acid (alpha-ABA) for 7 days. In microLCRA gamma mice alpha-ABA produced up to a 43-fold induction of gamma and 9-fold induction of mouse alpha globin genes. In contrast, butyrate did not induce gamma globin expression in the beta locus YAC mice. However, the gamma globin genes of beta locus YAC mice were activated after administration of 5-azacytidine (5-azaC), and the level of gamma globin expression was further increased by administration of alpha-ABA. These results suggest that butyrate cannot reactivate a totally silenced gamma gene and that induction of fetal hemoglobin by this compound may require the presence of preactivated gamma globin genes.