Organochlorine compounds (OC) are known to induce vitamin A (retinoids) deficiency in mammals, which may be associated with impairment of immunocompetence, reproduction and growth. This makes retinoids a potentially useful biomarker of organochlorine impact on marine mammals. However, use of retinoids as a biomarker requires knowledge about its intrapopulation patterns of variation in natural conditions, information which is not currently available. We investigated these patterns in a cetacean population living in an unpolluted environment. 100 harbour porpoises Phocoena phocoena from West Greenland were sampled during the 1995 hunting season. Sex, age, morphometrics, nutritive condition, and retinol (following saponification) and OC levels in blubber were determined for each individual. OC levels found were extremely low and therefore considered unlikely to affect the population adversely: mean blubber concentrations, expressed on an extractable basis, were 2.04 (SD = 1.1) ppm for PCBs and 2.76 (SD = 1.66) ppm for tDDT. The mean blubber retinol concentration for the overall population was 59.66 (SD = 45.26) mu g g(-1). Taking into account the high contribution of blubber to body mass, blubber constitutes a significant body site for retinoid deposition in harbour porpoises. Retinol concentrations did not differ significantly between geographical regions or sexes, but they did correlate significantly (p <0.001) with age. Body condition, measured by determining the lipid content of the blubber, did not have a significant effect on retinol levels but the individuals examined were considered to be in an overall good nutritive condition. It is concluded that measurement of retinol concentrations in blubber samples is feasible and has a potential for use as a biomarker of organochlorine exposure in cetaceans. However, in order to do so, biological information, particularly age, is critical for the correct assessment of physiological impact