Cold-storage preservation of the canine pancreas prior to islet isolation has previously been noted to reduce the intrasplenic islet autograft success rate; but the mechanism of this deleterious effect has not been determined. We undertook a study in both outbred dogs and Lewis (RT1-1) rats to determine the influence of cold-storage preservation interval, preservation solution, and flushing technique on islet yield and islet viability. The preservation solutions used were those that had proved most efficacious in preserving segmental canine pancreases--namely, the modifications of silica gel fractionated plasma (SGF-III and SGF-IV) and an hydroxyethylstarch/lactobionate solution (UW-1). In the first set of experiments, the traditional vascular flush was used; this was followed by storage at 4 degrees C. After brief periods of preservation (3 hr in the rat, 12 hr in the dog) there was a significant (P less than 0.006) reduction in islet yield. The reduced yields were similar with each solution tested, were made worse with increasing intervals of storage, and resulted in a significant reduction in autograft success rate. The second set of experiments examined the effect of using an intraductal flush prior to preservation, along with the effect of adding collagenase to the preservation fluid. Islet yields were maintained at control values in both animal models using preservation intervals of up to 24 hr. These islet yields produced auto- or isograft success rates similar to those obtained by transplanting freshly obtained tissue; verifying adequate islet viability. We recommend that a pre-storage ductal flush technique be used for cold-storage preservation of the pancreas prior to islet isolation and transplantation.