The play of a breeding colony of captive harbour seals (Phoca vitulina) was documented over a four-year period to identify changes associated with seasonal variations in energy use, and to corroborate previous field studies which were restricted to observations at the hauling grounds during breeding and moulting. Outside the breeding/moulting period, play was more frequent and in some animals average daily food intake was higher. Thyroid hormones did not vary seasonally in any clear manner. The adults increased their daily activity markedly during the breeding/moulting period. It is argued that these changes in play rate were a reflection of constraints on time rather than energy, since the animals played most when mass and fat were being accumulated for the winter. Field observations were confirmed in the laboratory, showing most play to be solitary, with significant amounts of adult play which was more stereotyped than that of younger seals. Behaviour in captivity was less ebullient and sometimes truncated in comparison to that seen in the wild; however, nine new types of play and a more elaborate repertoire of object play were seen in the laboratory, where seals could be observed throughout the year and underwater.