Density gradient separation of islets from exocrine tissue is usually performed with Ficoll. However, this reagent adds significantly to the cost of the isolation. The aim of this study was to evaluate the performance of Dextran as a potential low-cost substitute for Ficoll and to evaluate the effects of cold storage of the pancreatic digest prior to purification. Pancreases were procured from mongrel dogs, loaded with collagenase and mechanically dissociated. Washed pancreatic digest was collected and divided into two fractions that were purified using discontinuous gradients on the Cobe 2991 processor using identically prepared EuroFicoll (EF) or EuroDextran (ED) gradients. Alternate groups were suspended in EC and stored on ice, while the other fraction were resuspended in the 1.108-g/mL gradient layer (either EF or ED) and loaded into the COBE. This tissue layer was overlaid with layers of densities 1.096 and 1.037 g/mL and a HBSS cap, and centrifuged for 5 min at 800 x g. Purified islets were collected from the interface between the 1.037 and 1.096 layers and islet recovery, purity, and function were assessed. From a series of eight isolations, 72.9 +/- 8.2% (mean +/- SEM) of the islets were recovered from the EF purified gradients compared with 62.6 +/- 8.3% from ED gradients (p = NS, paired t-test). Gradients of ED that were run following hypothermic storage of the digest in cold EC solution (stored ED) had reduced islet recovery when compared with islet recovery from gradients prepared in EF(stored EF) (51.6 +/- 9.6% for ED stored vs. 72.9 +/- 11.9 for EF stored, p < 0.05). Islet recovery from EF gradients was equivalent between the stored and nonstored groups. The purity of preparations from the stored ED gradients was also reduced (71.3 +/- 4.3%) when compared with islets that were immediately purified after dissociation (82.5 +/- 4.8%, p < 0.05). Static glucose stimulation assay showed equivalence between the islets from ED and EF gradients. The stimulation index (SI) was 9.3 +/- 0.9 for EF islets compared with 7.9 +/- 1.4 for ED islets for digest purified immediately. However, if the digest was hypothermically stored in EC solution, a decrease in functional viability was observed in both the EF and the ED groups (7.7 +/- 1.4 and 5.9 +/- 0.8, respectively). Out of five alloxan-induced diabetic nude mice transplanted under the kidney capsule with 2000 islets isolated from the nonstored groups, three remained euglycemic >50 days posttransplant with either EF or ED islets. These experiments demonstrate effective recovery of equivalent numbers of canine islets using discontinuous gradients of ED or EF immediately following enzymatic digestion. However, following storage of the digest in cold EC solution results in a reduction in both islet recovery and function when gradients of ED are utilized.