Background: Treating neuropsychiatric problems in Parkinson's disease (PD) remains challenging. Only rivastigmine has US Food and Drug Administration approval for the treatment of PD dementia. Some studies have shown that atypical neuroleptics adequately improve hallucinations while others suggest that side effects and increased mortality negate benefit. We utilized a chart review to assess the tolerability and efficacy of commonly used agents for neuropsychiatric symptoms in PD. Methods: We reviewed patients with a PD diagnosis in our clinic database who were being treated with antidementia agents (donepezil, galantamine, rivastigmine, memantine) or atypical neuroleptics (quetiapine, olanzapine, clozapine) by two PD specialists over 13 years. We determined tolerability by therapy retention in comparison with the accepted gold standard, ie, rivastigmine for dementia and clozapine for hallucinations. Clinician documentation of patient/ caregiver or personally perceived clinical improvement was recorded to define therapeutic response. Secondary outcome measures were death, treatment-limiting side effects, and weight change. Results: A total of 171 PD patients met criteria for study inclusion. All medications were similarly retained in comparison with the standard, with the exception of donepezil which was significantly better (P , 0.02) than rivastigmine. Using the documentation metric, no medication was statistically superior to its standard. Olanzapine approached significance because it was inferior to the clozapine standard (P , 0.08). Secondary outcome measures were similar among medications. Conclusion: We conclude that, as a group, acetylcholinesterase inhibitors, atypical neuroleptics, and memantine are well tolerated in treating either cognitive difficulties or hallucinations in PD. They appear to offer variable levels of clinical benefit, as well per documented therapeutic responses. Side effects, death, and weight change were not significantly different among medications.