In two lines of White Leghorns that differ in their propensity to feather peck, feather pecking was already present I day after hatching. A significant line difference developed within 3 days and this difference remained in the following weeks. There was no clear relationship between feather pecking and ground pecking, instead feather pecking was associated with other socially orientated pecks. When chicks were confronted with feathers as an inanimate stimulus pecking at these feathers did not reflect pecking in a social context. These results suggest that feather pecking has an underlying social component, rather than being a redirected behaviour. To test the influence of social factors on feather pecking, chicks from the high feather pecking line were subjected to regular rehousing or regular rehousing plus confrontation with unfamiliar peers. Gentle feather pecking increased significantly in,or, cups that were rehoused with unfamiliar peers. Directly after confrontation chicks displayed a preference in the total number of pecks toward these unfamiliar birds. Gentle feather pecking as a separate pecking orientation showed a similar trend. Increased righting times in a tonic immobility test were recorded in the experimental groups, hence frequent confrontation with unfamiliar peers may be experienced by chicks as stressful, which is consistent with the hypothesis that stress mediates the expression of feather pecking. In contrast to the dominant hypotheses in the field, we argue that gentle feather pecking at an early age plays an important role in the building (social exploration) and maintenance of social relationships between chicks. (C) 2002 Elsevier Science B.V. All rights reserved.