Playing with toys as a means of environmental engagement has long been considered important in early child development (Messer, Rachford, McCarthy, & Yarrow, 1987; Wolery & Werts, 1994). However, children with highly significant disabilities often engage in toy play less frequently than their peers without disabilities (Blasco, Bailey, & Burchinal, 1993; Langley, 1985) and frequently need specialized support to promote toy play (Ivory & McCollum, 1999; Langley, 1985). Preference assessment technology was used to identify preferred sensory attributes. Those sensory attributes were embedded into a toy previously identified as nonpreferred based on selection and amount of toy play. Differences were noted in caregiver opinion of sensory preference when compared to a sensory attribute preference assessment. The initial paired-item presentation consisted of a rotation of 4 toys from the natural environment using a child-directed approach, which consisted of choice, prompting, and praise for completion of functional toy play. An additional condition added the child's preferred sensory attribute to a nonpreferred toy using the above-mentioned procedures. Results indicated that the addition of a preferred sensory attribute increased selection of that toy with 2 of 3 children, and resulted in increased independent functional toy play for all 3 children.