In this study, the ob/ob mouse model was used to investigate epidemiological evidence linking fish intake to relative reduction in incidence of Type 2 diabetes mellitus and glucose. We have investigated, in comparison to low and high fat diets, the effect of a fish oil diet on basal and stimulated plasma glucose and insulin levels in male and female ob/ob mice. Mice were fed for 12 months with a saturated fat diet containing 25% lard, with a low fat diet containing 5% soybean oil, with a polyunsaturated fat diet containing 25% safflower seed oil (n-6) or with polyunsaturated fat diet containing 23% fish oil (n-3). Total body weight increased to approximately 100 g at the end of the experiment, with the highest increase in the order of lard > safflower oil > fish oil > soybean oil diet. Intercurrent deaths were found especially in the fish oil diet group. Compared to the other diet groups, plasma insulin levels of the fish oil diet group were significantly increased 3 months after the start of the diet and remained higher for another 3 months. Thereafter, the level declined to those of the other diet groups. Glucose-tolerance tests at 3, 6, 8 and 10 months showed a tendency of more efficient tissue glucose uptake in the fish oil group compared to the other groups, which was in accordance with a higher plasma insulin levels. At 12 months, microscopy revealed an increased severity of hepatic brown pigment accumulation and extramedullary haematopoiesis in the spleen of mice fed with fish oil. We conclude that fish oil diet in ob/ob mice reduced the body weight gain and increased the glucose-induced insulin secretion. Fish oil diet also increased intercurrent mortality. However, a consistent course of death could not be established using morphological parameters.