We have identified two novel human genes encoding proteins with a high level of sequence identity to two previously characterized RNA-binding proteins, αCP-1 and αCP-2. Both of these novel genes, αCP-3 and αCP-4, are predicted to encode proteins with triplicated KH domains. The number and organization of the KH domains, their sequences, and the sequences of the contiguous regions are conserved among all four αCP proteins. The common evolutionary origin of these proteins is substantiated by conservation of exon–intron organization in the corresponding genes. The map positions of αCP-1 and αCP-2 (previously reported) and those of αCP-3 and αCP-4 (present report) reveal that the four αCP loci are dispersed in the human genome; αCP-3 and αCP–4 mapped to 21q22.3 and 3p21, and the respective mouse orthologues mapped to syntenic regions of the mouse genome, 10B5 and 9F1-F2, respectively. Two additional loci in the human genome were identified as αCP-2 processed pseudogenes (PCBP2P1, 21q22.3, and PCBP2P2, 8q21–q22). Although the overall levels of αCP-3 and αCP-4 mRNAs are substantially lower than those of αCP-1 and αCP-2, transcripts of αCP-3 and αCP-4 were found in all mouse tissues tested. These data establish a new subfamily of genes predicted to encode closely related KH-containing RNA-binding proteins with potential functions in posttranscriptional controls.