<p>For people with dementia, meaningful engagement with preferred activities is a vital component in maintaining quality of life. Preference assessments have been found to identify preferred and reinforcing stimuli for people with dementia. However, because they have been predominantly validated for individuals with intellectual or developmental disabilities, there are likely aspects of the assessment that need to be adapted. We implemented standard stimulus preference assessments for four people with dementia, and designed some adaptations based on objective measures of their choice, affect, verbal responding, and the input of the participants. We then compared the standard and the adapted versions in a multielement design. We found that both the standard and adapted assessments resulted in similar selections of stimuli. We observed an increase in indices of happiness and a decrease in indices of unhappiness for three out of four participants during the adapted version compared to the standard version. Family, staff, and the people with dementia reported both the standard and adapted preference assessments to be socially valid. One participant reported better understanding what they were being asked to do in the adapted assessment, and a staff member reported the adapted assessment to be more effective. We also found a decrease in the frequency of statements that were off-topic, socially polite, or about items being “sold” to them across all four participants in the adapted assessment. We further extended the work of Reid and Green (1996) by incorporating indices of confusion to capture a broader range of affect. On the basis of our results, we give some examples of how to make preference assessments more acceptable for people with dementia.</p>