The acute effects of anoxia are dramatic in animals that have high standard metabolic rates. Mammals, for instance, suffer a loss of brain functions within seconds, whereas the behaviour of some lower vertebrates such as frogs appears little affected in the initial phase of anoxia so that their reactions to anoxia can be studied on an extended timescale. Adult insects have very high metabolic rates and during anoxia all organs rapidly cease to function. In contrast to mammals, however, which will suffer irreversible damage if anoxia is extended for more than a few minutes, many insects can survive anoxia for several hours in a state of rigor and fully recover when oxygen is readmitted. This makes insects apt experimental animals for studying the question as to which structural, physiological and metabolic properties improve the tolerance of anoxia. We have followed the heat dissipation (as an index of metabolic rate) during anoxia and postanoxic recovery in frogs and in two species of insects, a locust and a moth. Frogs can survive 3-4 h of anoxia (at 20-degrees-C), locusts 4-5 h, whereas the moth Manduca sexta can fully recover from more than 20 h of strict anoxia. In frogs heat flow was decreased to about 25% during 90 min of anoxia, but in anoxic insects heat dissipation was reduced much further, to about 5% in locust (during 4 h) and to less than 4% in Manduca (during 14.5 h of anoxia). When air was readmitted heat, dissipation, was rapidly increased to a level exceeding the normoxic control value (overshoot). Locust and moth differed strikingly with respect to heat dissipation during postanoxic recovery. Locusts dissipated a large amount of excess heat during, recovery for more than 12 h, in contrast to Manduca where excess heat during recovery was small and the control value was reached within less than 2 h. Metabolic processes underlying the different patterns of heat dissipation during anoxia and postanoxic recovery are discussed.