Parkinson's disease (PD) has long been thought of as a debilitating motor disease: only relatively recently has research focused on cognitive functioning. It is now widely accepted that memory processes are among the primary cognitive functions to deteriorate in PD. However, less is known about the role that task variables (e.g., difficulty) and participant characteristics (e.g., gender, disease progression) play in this relationship. In addition, few studies in the PD literature have looked at the important issues of statistical power and the magnitude of memory deficits. The present investigation addressed some of these concerns. The first stage involved conducting a power analysis, based on 48 studies, followed by a meta-analysis. The meta-analysis included 32 effect sizes from studies assessing recognition memory in PD. This analysis paved the way for a large-scale study examining recognition memory in 41 nondemented PD participants compared to 41 age- and education-matched healthy controls. Both verbal and nonverbal recognition tasks were specifically designed for this purpose, the latter employing two levels of difficulty. In addition, prospective memory (remembering to remember) was examined with two event-based tasks because no study to date has looked at this issue in PD. The results of the power analysis indicated that past research has typically had insufficient statistical power to detect all but the largest memory deficits. Integrating the data from many studies, the meta-analysis suggested that nondemented, medicated PD participants may suffer from small recognition deficits. Support was provided by the subsequent primary study. In addition, it was found that the progression of memory impairment operates in parallel with the progression of motor symptoms. Moreover, task demands interacted with disease stage, such that nonverbal recognition deficits were only seen in early-stage PD participants when the task was made more difficult. Conversely, advanced-stage participants produced deficits irrespective of the level of difficulty. With respect to prospective memory, only the advanced-stage participants showed clear evidence of memory impairment. The main outcome of this research is that recognition memory impairment does occur in PD. However, low levels of statistical power in previous research and moderating factors such as symptom severity and task difficulty have likely obscured true deficits.