Serial blood samples were collected from the unfed spiny dogfish shark (Squalus acanthias) by use of a cannula chronically implanted in the dorsal aorta. Plasma glucose, alanine, lactate, and beta-hydroxybutyrate levels were estimated by standard enzymatic procedures. The metabolite levels were estimated immediately after the surgery, for an additional 3 to 5 postoperative days before experimental use, and for 7 to 9 days after a single infusion of mammalian insulin (50, 100, and 250 IU/kg body wt) or the control solution. Plasma glucose, alanine, and lactate levels declined for 1 or 2 days after the surgery, but beta-hydroxybutyrate levels were essentially unchanged. Plasma glucose levels remained relatively stable for the remainder of the pretreatment period, and in the control animals during the subsequent treatment period. In contrast, plasma alanine and beta-hydroxybutyrate levels increased during both the pretreatment period and the treatment period in the control animals. Plasma lactate levels increased in the control animals after Day 3 of the treatment period. Insulin infusions resulted in severe, prolonged depressions of plasma glucose and alanine levels. Plasma lactate levels also fell after insulin infusion, but the depressed levels were interrupted by precipitous, large increases, followed by sharp declines, that were evident for one to three sampling periods and whose onset varied between Days 3 and 5 in individual animals. The levels increased again in some animals between Days 6 and 9. Plasma beta-hydroxybutyrate levels were comparatively resistant to insulin. After an early decline, plasma beta-hydroxybutyrate rose to levels that were near the control levels by Days 2 through 4, but fell to near the initial levels by Day 5 and thereafter. The results suggest that the spiny dogfish, a species that consumes reliable and abundant amounts of protein and lipid, but little carbohydrate, and that may feed as seldom as every 2 weeks, uses ketone bodies (beta-hydroxybutyrate and presumably acetoacetate) as primary fuels. Glucose is apparently synthesized by gluconeogenesis, alanine and lactate are probable substrates, and ketone bodies are the likely sources of energy. The evidence supports the hypothesis that glucose is synthesized to maintain muscle glycogen reserves that are used for anaerobic glycolysis when the animal is engaged in predation, escape, or other circumstances that dictate vigorous swimming. If other tissues and organs require glucose as a fuel source, the amounts must be exceptionally small.