At around day 11 of life, chicks show a tendency to move out of sight of their mother before returning and regaining social and visual contact. A series of experiments was conducted to investigate the role of this voluntary 'out-of-sight' behaviour on the development of spatial memory in young chicks. Performance on various spatial tests was compared between chicks reared in environments providing the opportunity to move out-of-sight of an imprinting object (occlusion-experienced chicks) and occlusion-naive chicks. As in natural conditions, a peak in out-of-sight behaviour was found on day 11 compared with days 10 or 12 (ANOVA, F2,20=20.2, P<0.001). Occlusion-experienced chicks walked more than occlusion-naive chicks when released into a large novel arena? (ANOVA, F1,14 = 11.9, P<0.01), but otherwise showed similar degrees of dispersal. Occlusion-experienced chicks tended to show better retrieval of a visually displaced imprinting stimulus than occlusion-naive chicks (Chi-squared test, '2=6.3,df=3 P=0.09). Occlusion-experienced chicks tended to make fewer orientation errors in the first (Kruskal-Wallis, '2=7.0, df=3, P<0.07) and subsequent trials (Kruskal-Wallis, '2=7.6, df=3, P=0.05) of the retrieval test. Chicks making at least one orientation error had spent less time out-of-sight of the imprinting stimulus on day 11 than chicks making no orientation errors (ANOVA, F1,18= 6.5, P<0.05). In contrast, experimentally manipulating the amount time that chicks were out-of-sight of an imprinting stimulus (by confining the chicks) had no effect on performance in displacement or detour tests. Results suggest that egocentric orientation in chicks is improved by allowing them to express out-of-sight behaviour in early life. The implications of providing appropriate stimulation during early life in ameliorating crowding and hysteria in large commercially-housed flocks is discussed.