At 21 days of age, 16 pairs of male laboratory mice of the ICR strain were weaned and allocated to four treatment groups in a 2×2 factorial design matched for genetic background (litter) and body weight. Factor one was the hardness of the food pellets with a significant 2.5-fold difference between soft and hard feed. Factor two was the environment, with half of the mice being kept in barren standard cages, while the other half were additionally provided with a cardboard tube. Subjects were videotaped during the full 12-h dark period on three occasions: 3 days after weaning, when stereotypies start to develop (24 days), at an early stage of stereotypy development (34 days), and when adult with fully established stereotypies (80 days). Since feed hardness had no effect on time spent feeding, the absence of an effect of the feeding treatment on stereotypic wire-gnawing remains inconclusive with respect to the role of feeding motivation in the development of this stereotypy. The interaction between the development of feeding and wire-gnawing, respectively, does not, however, suggest a strong relationship. In contrast, enrichment significantly reduced stereotypic wire-gnawing in adults by 40% ( F=4.47, df=1,26, p<0.05), presumably as a consequence of the cover provided by the cardboard tubes. This is substantiated by observations that the tubes were used as a place to retreat upon disturbance as well as for resting. As a consequence, when adult the mice showed more resting ( F=6.46, df=1,26, p<0.05) and less grooming ( F=9.79, df=1,26, p<0.01) in cages containing them, suggesting that mice with access to a cardboard tube (and hence shelter) perceived a greater level of security in these cages. Cardboard tubes therefore provide a simple means of environmental enrichment to reduce the development of abnormal behaviours in laboratory mice and may thus be beneficial in terms of their well-being.