A precursor to the gastrointestinal hormone secretin has been isolated. The starting material for the purification of the precursor was a peptide fraction purified from pig intestinal extracts, containing peptides with a molecular weight higher than that of secretin. The purification could be followed by measurement of secretin bioactivity (alkali secreted in the pancreatic juice of anesthetized cat). Sequence analysis of the isolated secretin precursor revealed a 71-amino acid residue polypeptide that contained the sequence of secretin N terminally, followed by a Gly-Lys-Arg sequence and a C-terminal extension of 41-amino acid residues. With the exception of an arginine residue, which occurs directly after the Gly-Lys-Arg sequence, the remainder of the C-terminal residues in this precursor are identical to the 40 C-terminal residues predicted by the recently described cDNA sequence for porcine preprosecretin. Compared to the deduced preprosecretin sequence, a stretch of 32 amino acid residues directly following the Gly-Lys-Arg sequence is missing in the now purified secretin precursor. This implies that differential splicing may occur when the secretin gene transcript is processed to mRNA.