Background. While the efficacy of cognitive stimulation (CS) has been demonstrated in patients with dementia, no study has included patients with Parkinson's disease dementia (PDD). Objective. For the first time, this randomized crossover pilot study examined the feasibility and potential effects of CS in PDD. Methods. All residents of a PDD-specific long-term care unit in the Netherlands that were eligible for the study (n=12) were randomly allocated to group A (n=6) receiving CS (eight weeks, twice weekly for 60minutes) or group B (n=6) receiving usual care (control group, CG). The CG participated in CS afterwards, resulting in an experimental group (EG), consisting of n=12. Pre- and postassessments and a six-week follow-up (FU) were conducted for cognition, neuropsychiatric symptoms, quality of life (QoL), and activities of daily living (ADL) outcomes. Results. Between-group analysis with difference scores from pre- to posttest revealed a group difference for global cognition (CERAD total score) favoring the EG, with a moderate effect size and a p value just failing to reach statistical significance (p=0.067; r = 0.43). A further statistical trend was observed for neuropsychiatric symptoms, again with a moderate effect size (p=0.075; r = 0.42). Within-group analyses indicated improvement only in the EG with large effects also just failing to reach significance for global cognition (short term, p=0.060; r = 0.70) as well as for depression (long term, p=0.072; r = 0.61). ADL deteriorated significantly at FU in the EG (p=0.014; r = 0.71). Conclusions. Although our data are preliminary due to the small sample size, this study shows that CS is feasible and potentially effective for cognitive and noncognitive outcomes in PDD patients. Randomized controlled trials with larger sample sizes are needed to confirm these promising results.