The aim in this presentation is to review available data concerning the provision of enrichment to kennelled dogs, and to illustrate some of the conceptual issues requiring resolution before composite enrichment programs should be universally recommended. This is important because growing awareness of the implications of animal use and housing on welfare have led to rising public expectations and lower levels of tolerance for conditions not perceived to be adequate. Consequently, there have been calls for improved provision for the welfare of animals housed in captivity, including the use of enrichment programs. Enrichment can be broadly understood as the provision of a more complex social, physical and sensory environment, with the goal of improving the welfare of the target animals. Various elements of enrichment have been recognised as having a significant positive effect on the behaviour, physiology and welfare of animals housed in captivity, including kennelled dogs. However, composite programs containing multiple elements of visual, auditory, olfactory and social enrichment risk overstimulating these dogs, and may increase levels of stress, reducing rather than improving, levels of canine welfare. This is potentially problematic, not only to the welfare of the animals, but also their ability to provide the desired service for humans. If too much stimulation leads to increased stress, potential working dogs may fail the training and accreditation programs for substance detection, guide or assistance work. Failing to acknowledge that a composite enrichment program's effects may not equal the sum of its parts also risks the investment of more resources than is actually required. There is clearly a need for systematic research, conducted in realistic industry conditions, to evaluate perceptions regarding the importance of enrichment for the welfare of kennelled dogs and to determine whether composite enrichment programs affect the welfare, training success and resource requirements of such dogs.