To investigate the patterns and mechanisms of capture-induced hyperthermia, we surgically implanted 26 impala (Aepyceros melampus) with miniature thermometric data loggers, which measured body temperatures continuously throughout capture procedures. Four groups of impala, which were habituated to varying levels of handling and boma-housing, were captured by net restraint or by chemical immobilization. The study took place between July 1999 and December 2005. Irrespective of whether impala were chemically captured, net-captured, or disturbed by exposure to a stressor, they developed a precipitous increase in body temperature. This increase in body temperature was not related to activity levels; animals that had low activity levels before immobilization had larger increases in body temperature compared to those that had high activity levels but were not immobilized (t=3.6, P=0.001, n=5). Similarly this increase in body temperature was not related to environmental heat load at the time of darting and immobilization (r=−0.05, P=0.85). Body temperature increase also did not depend on whether the animals were captured using drugs or not. However, we found that those animals that were habituated more to handling and boma-housing had smaller increases in body temperatures (F=37, P<0.001) and smaller stress responses, indicated by lower plasma cortisol concentrations (F=5.5, P<0.05), and less fractious behavior, compared to those animals that were habituated less or not at all. Therefore we believe that capture-induced hyperthermia in impala is caused predominantly by stress, which induces a rapid rise in body temperature.