By the use of hybrids between a U1 small nuclear ribonucleoprotein (snRNP: U1A) and a U2 snRNP (U2B") we have identified regions containing 29 U1A-specific amino acid residues scattered throughout the 117 N-terminal residues of the protein, which are involved in binding to U1 RNA. The U1A-specific amino acid residues have been arbitrarily divided into seven contiguous groups. None of these groups is sufficient for U1 binding when transferred singly into the U2B" context, and none of the groups is essential for U1 binding in U1A. Several different combinations of two or more groups can, however, confer the ability to bind U1 RNA to U2B", suggesting that most or all of the U1A-specific amino acid residues contribute incrementally to the strength of the specific binding interaction. Further evidence for the importance of the U1A-specific amino acid residues, some of which lie outside the region previously shown to be sufficient for U1 RNA binding, is obtained by comparison of the sequence of human and Xenopus laevis U1A cDNAs. These are extremely similar (94.4% identical) between amino acid residues 7 and 114 but much less conserved immediately upstream and downstream from this region.