Further to the well-established association between apolipoprotein E (APOE) and Alzheimer's disease, this gene has also been implicated in both susceptibility to, and dementia in, Parkinson's disease (PD). However studies to date have produced contradictory findings. We conducted a case-control study in a population of 528 PD patients and 512 healthy controls and found no significant difference in allele or genotype distribution of APOE between the two groups. An updated meta-analysis showed a modest increase of APOE-epsilon2 carriers amongst PD patients compared to controls [P = 0.017, OR = 1.16 (95 % CI 1.03-1.31)]. 107 of our patients were incident cases participating in a population-based epidemiological study. Longitudinal follow-up of this cohort over a mean of 5.0 +/- 0.7 years from diagnosis revealed no significant impact of APOE-epsilon4 carrier status on risk of dementia or rate of cognitive decline. An updated meta-analysis indicated an over-representation of APOE-epsilon4 carriers amongst PD dementia compared to non dementia cases [OR 1.74 (1.36-2.23), P = 1 x 10(-4)], although small sample sizes, heterogeneity of odds ratios and publication bias may have confounded this finding. In conclusion, our study does not support previously reported associations between APOE genotype and susceptibility to, or cognitive decline in, PD. An updated meta-analysis indicates any association with PD susceptibility is at most modest, an observation with important implications for further study of this issue. Large scale longitudinal studies would be best placed to further evaluate any impact of APOE genotype on cognitive decline in PD.