A previously described VPAC2-selective agonist, BAY 55-9837 (peptide HSDAVFTDNYTRLRKQVAAKKYLQSIKNKRY), had several limitations with respect to its potential as an insulin secretagogue for the treatment of type 2 diabetes. These limitations were primarily poor stability in aqueous buffer and short duration of action in vivo. In this report, we describe a series of novel analogs of BAY 55-9837 that were designed around the likely degradation mechanisms and structure-activity relationship of this peptide with a view to overcoming its limitations. These analogs were tested for improved liquid stability and retention of VPAC2-selective binding and activation, as well as prolonged activity in vivo. Although several degradation mechanisms were possible based on the degradation pattern, it was determined that deamidation at the two asparagines (N9 and N28) was the major instability determinant. Changing these two asparagines to glutamines did not negatively affect VPAC2-selective binding and activation. The double glutamine mutein analog, BAY(Q9Q28), retained full VPAC2 activity and selectivity while displaying no significant degradation when stored at 40 degrees C for 4 weeks. This is in contrast to BAY 55-9837, which showed greater than 80% degradation when stored at 40 degrees C for 2 weeks. A cysteine was added to the C terminus of BAY(Q9Q28), followed by site-specific cysteine conjugation with a 22- or 43-kDa polyethylene glycol (PEG) to yield BAY(Q9Q28C32)PEG22 or BAY(Q9Q28C32)PEG43, respectively. These PEGylated peptides retain the ability to selectively bind and activate the VPAC2 receptor and have prolonged glucose-lowering activity in vivo.