The objective of this study was to determine the effects race, sex and sleep disturbance have on the severity of Parkinson’s disease as assessed by the Hoehn Yahr (HY) score in both the medicated (ON) and non-medicated (OFF) states. The potentially confounding variables of age, time in years from the onset of symptoms to database entry, and education were taken into account. Secondary analysis was also conducted to determine how the non-motor symptoms of dementia, hallucinations and autonomic dysfunction impacted Hoehn Yahr ON and OFF scores. This study used the statistical techniques of the Student’s t-test, ANOVA, Tukey-Kramer test, univariate linear regression, and multivariate regression. The t-tests and ANOVA test revealed that there was no significant differences in mean HY ON and OFF scores between the sexes, patients with and without sleep disturbance, and between the different races analyzed in this study. Patients with and without sleep disturbance did show significantly higher HY ON scores as compared to HY OFF scores, which is peculiar as this finding suggests that these patients are not responding to their medication. The univariate linear regression models did show, however, that time in years from the onset of symptoms to database entry did significantly impact both HY ON and OFF scores, whereas age is only shown to have a significant impact on HY OFF scores. Additionally, the univariate linear regression model analyzing the association between education and HY OFF scores showed that having some high school education, but not receiving a degree, was associated with an increase in HY OFF scores. Several multivariate linear regression models where built to assess the impact different predictor variables had on HY ON and HY OFF scores. The first two multivariate models used the predictor variables of age, race, and time in years from the onset of symptoms until database entry. These models showed that only time in years from the onset of symptoms until database entry impacted HY ON scores, whereas all three of these predictor variables impacted HY OFF scores. Two additional multivariate linear regression models were built to assess how age, race, time in years from the onset of symptoms until database entry, dementia, autonomic dysfunction and hallucinations all impacted HY ON and OFF scores. These models revealed that all of these predictors, when taken together, significantly impacted HY OFF scores, but not HY ON scores. Finally a scatter plot was made comparing HY ON and HY OFF scores. A LOWESS scatter plot smooth line was also superimposed on top of this plot to show the overall trend these scores had on one another. This scatter plot was interesting because it suggested that there were two spate groups of patients contained in this database, those that responded well to medication and those that did not. Overall, this study showed that age, time in years from the onset of symptoms until database entry, education and race impacted HY OFF scores. Furthermore, the analysis indicated that patients who were asked about sleep disturbance did not appear to be responding to medication. There are several limitations to this study, however, with the most important being missing data and the cross-sectional design. Missing data prevented sleep disturbance from being thoroughly analyzed and the cross-sectional design does not allow for any causal relationships to be determined.