The effect of previous experience on the behaviour of gilts was investigated. Twenty gilts were reared, from birth to farrowing, in an enriched environment and 20 in a barren environment. Six weeks prior to the expected date of parturition the gilts underwent a handling test and an open field test. A few days prior to parturition the gilts were moved into farrowing accommodation. There were two farrowing environments, barren and enriched, and, using a crossover design, gilts were moved either to a similar environment to that which they had previously experienced, i.e. barren to barren (BE) or enriched to enriched (EE) or to a different environment, i.e. barren to enriched (BE) or enriched to barren (EB). Gilts from enriched environments were more difficult to drive, scored on a scale of 1-5 (B 1.2, E 1.8, SEM 0.18, P < 0.05) and vocalized more frequently in the 7 min test period (B 32.4, E 69.5, SEM 10.46, P(0.05) than gilts from barren environments while the latter gilts performed more locomotory behaviour in the open field test than their counterparts from enriched environments (B 10.6, E 8.9, SEM 0.44, P < 0.05). Gilts which experienced a change of environment at parturition exhibited more active behaviour than gilts which were moved to farrowing accommodation similar to the housing in which they were reared (BE 2%(a), BE 28(b), EE 7(a), EB 27(b), SEM 5.1, P < 0.01). The influence of the environment in determining pig behaviour is discussed with reference to the effects of past experience.