Metabolic cages are used for housing rats and mice for up to five days for collection of urine and/or faeces. The small, barren area of the metabolic cage compromises animal welfare as the animals lack a solid floor, shelter, nest material and social contact. We constructed and tested a practically-applicable enrichment device designed to meet behavioural needs for environmental complexity. The influence of this device on the cage preferences and stress levels of the animals was evaluated. A box-shaped enrichment device was designed and implemented in existing metabolic cages. Male Tac:SD rats were housed for five days in an enriched metabolic cage (EMC; n = 12) or a standard metabolic cage (SMC; n = 12), and data were collected on bodyweight, food and water intake, urination and defaecation, as well as urinary corticosterone and creatinine. Moreover, open-field behaviour and cage preferences were assessed. Rats in both groups gained significantly less weight when housed in metabolic cages. Furthermore, SMC rats failed to increase their weight gain after being housed in the metabolic cage. Defaecation was significantly higher in the SMC than in the EMC and so was urinary creatinine. No group differences were found in open-field behaviour. However, in comparing activity before and after housing in the metabolic cage, only SMC animals exhibited significantly lower total activity. In a preference test, a preference for the tunnel connecting the cages in the preference test and a side preference for the left side were found. This side preference was eliminated when the EMC was placed on the right side, whereas the right side was significantly avoided when the EMC was placed on the left side. Based on these results, we conclude that, to some extent, the enrichment device improved the welfare of rats housed in EMC, compared to those in SMC.