Engagement is the active involvement in experiences which allows for development. For children with severe mobility impairments however, challenges arise in accessing experience which can lead to learned helplessness. Due to these challenges powered mobility has been suggested as a mechanism for the provision of self-initiated access to experiences. However, powered mobility is out of reach of the majority of children with disabilities in South Africa hence a non-powered alternative has been sought. This study sought to determine the effect of non-powered, self-initiated mobility on the engagement of young children, with severe mobility impairment, in play. A multiple probe design across participants was used. Four participants, aged 2 years 10 months to 6 years 9 months with severe mobility impairment (Gross Motor Function Classification Scale (GMFCS) level V) underwent a minimum of 5 baseline sessions, followed by 8 intervention sessions. Engagement was measured during each session using the Individual Child Engagement Record –Revised (ICER-R). The data was analysed using visual graphic and statistical analysis. All participants demonstrated an improvement in engagement in play with the introduction of non-powered, self-initiated mobility. A reciprocal deterioration in non-engagement was also recorded. A decrease in engagement in play at the start of intervention was attributed to the focus of engagement being on mobility skills as the device was introduced but this reverted as the participants spent more time on the mobility device. Functional abilities were identified as having a greater role in engagement than age. The introduction of non-powered, self-initiated mobility correlated with the improvement of engagement of young children with severe motor impairment. Non-powered, self-initiated mobility is a viable, cost effective mechanism for mobility at a young age.