Before clinical islet transplantation can become an effective and reliable treatment for type 1 diabetic patients, there must be significant improvements in the methods employed for the isolation of islets of Langerhans. We have developed an automated cell extraction system (ACES), which allows computer control of the isolation process. As well, it incorporates a novel method of recombining dissociated pancreatic tissue. Following initial system design and testing to determine the optimal system configuration, a series of 12 consecutive canine islet isolations were performed. Pancreases were perfused with collagenase via the duct and dissociated and recombined using either the standard Ricordi-based protocol (group 1, n = 6) or dissociated and recombined using the ACES system (group 2, n = 6). A total of 90.8 +/- 21 x 10(3) islet equivalents (IE) (mean +/- SEM) were recovered in group 1 vs. 99 +/- 14 x 10(3) IE in group 2 (p = NS, student unpaired t-test). Following Ficoll purification the recovery was 56.2 +/- 14 x 10(3) IE for group 1 vs. 54.7 +/- 11 x 10(3) IE for group 2 (p = NS). Viability was equivalent with an 8.6-fold increase in insulin secretion for group 1 and an 8.8-fold increase for group 2 when the islets were exposed to high glucose solution supplemented with IBMX (3-isobutyl-1-methylxanthine) during static incubation. In vivo function was equivalent following transplantation of 2000 IE under the kidney capsule of alloxan-induced diabetic nude mice with five of six and five of seven mice surviving long-term (> 50 days posttransplant) (groups 1 and 2, respectively). This data shows that an entirely automated pancreatic islet extraction system can result in effective canine islet recovery without compromising islet yields and viability. The ACES system has several advantages over the standard isolation protocol. These include: 1) computer control and monitoring over all phases of the isolation, 2) a single-use sterile disposable tubing set, and 3) a novel method of tissue recombination.