Investigations in Prince William Sound (Alaska, USA) following the Exxon Valdez oil spill (EVOS) revealed that river otters (Lontra canadensis) on oiled shores had lower body mass and elevated values of biomarkers, than did otters living on nonoiled shores. In addition, otters from oiled areas selected different habitats, had larger home ranges, and less diverse diets than animals living in nonoiled areas. These differences between river otters from oiled shores and those from nonoiled areas strongly suggested that oil contamination had an effect on physiological and behavioral responses of otters. In this study, we explored the effects of crude oil contamination on river otters experimentally. We hypothesized that exposure to oil would result in elevated values of biomarkers, indicating induced physiological stress. Fifteen wild-caught male river otters were exposed to two levels of weathered crude oil (i.e., control, 5 ppm/day/kg body mass, and 50 ppm/day/kg body mass) under controlled conditions in captivity at the Alaska Sealife Center in Seward (Alaska, USA). Responses of captive river otters to oil ingestion provided mixed results in relation to our hypotheses. Although hemoglobin (Hb, and associated red blood cells) and white blood cells, and possibly interleukin-6 immunoreactive responded in the expected manner, other parameters did not. Aspartate aminotransferase, alanine aminotransferase, and haptoglobin (Hp), did not increase in response to oiling or decreased during rehabilitation. Conversely, principle-component analysis identified values of alkaline phosphatase as responding to oil ingestion in river otters. Our results suggested that opposing processes were concurring in the oiled otters. Elevated production of Hp in response to tissue damage by hydrocarbons likely occurred at the same time with increased removal of Hp-Hb complex from the serum, producing an undetermined pattern in the secretion of Hp. Thus, the use of individual biomarkers as indicators of exposure to pollutants may lead to erroneous conclusions because interactions in vivo can be complicated and act in opposite directions. Additionally, the biomarkers used in investigating effects of oiling on live animals usually are related to the heme molecule. Because of the opposing processes that may occur within an animal, data from a suite of heme-related biomarkers may produce results that are difficult to interpret. Therefore, we advocate the exploration and development of other biomarkers that will be independent from the heme cycle and provide additional information to the effect of oiling on live mammals.