We have previously reported on the identification of a cDNA clone encoding a novel human growth factor, named "CRIPTO," that is abundantly expressed in undifferentiated human NTERA-2 clone D1 (NT2/D1) and mouse (F9) teratocarcinoma cells. We now report the organization and nucleotide sequence of two related genomic sequences. One (CR-1) corresponds to the structural gene encoding the human CRIPTO protein expressed in the undifferentiated human teratocarcinoma cells, and the other (CR-3) corresponds to a complete copy of the mRNA containing seven base substitutions in the coding region representing both silent and replacement substitutions. The 440 bp 5' to the CAP site of CR-1 are preserved in CR-3. CR-1 maps to chromosome 3, and CR-3 maps to Xq21-q22. Southern blot analysis reveals that multiple CRIPTO-related DNA sequences are present in the human as well as in the mouse genome.