Utilizing the intrasplenic canine islet autograft model, it was possible to examine the effect of cold-storage of pancreatic tissue, both prior to and following dispersion, on functional outcome. A control group of dogs receiving freshly prepared autografts (n = 8) obtained durable euglycaemia in 75% of cases; the mean K value at two weeks being -1.70 (Standard Deviation (sd) = 0.06). In the first experimental group, animals (n = 8) were transplanted following 24 hours of whole pancreas cold-storage in silica gel fractionated plasma (SGF) using intraductal perfusion prior to preservation. None of these animals obtained euglycaemia; the transplanted material having a significantly reduced insulin content compared with controls (p = 0.0006). In the second experimental group, animals (n = 6) were transplanted following 24 hours of dispersed pancreatic tissue storage in SGF. This resulted in an 83% incidence of durable euglycaemia; the mean K value at two weeks being -1.60 (sd = 0.3). The glucose decay curve improved with time in this group and at three months the mean K value was -2.00 (sd = 0.38); this value being significantly superior to that of the control animals (K = -1.34, sd = 0.34, p less than 0.05). We conclude that while storage of the whole pancreas prior to islet isolation remains problematic, it is possible to reliably preserve dispersed pancreatic tissue for 24 hours by simple cold-storage, as assessed by the functional outcome of intrasplenic autografting in the dog model. These findings have important clinical implications.