Parkinson’s disease (PD) is a neurological disorder that nearly affects 1 million people in the US alone. PD causes involuntary motor fluctuations, resulting in bradykinesia (slowness of movement), tremors, stiffness, walking imbalance, and other complications. PD has no known cure; however, treatment exists through regular medication intake and therapy. Clinicians rate the progression of the symptoms based on standard clinical assessments during in-clinic visits that occur 2-3 times a year. In addition to self-reports from the patients, neurologists find it helpful if motor symptoms can be monitored objectively at in-home settings for better insights into disease progression. This work explores the impact of medication on motor symptoms, such as the spatial-temporal and frequency features of the finger-tapping motor exam. In our prior research, people with Parkinson’s Disease (n=4) were given to wear a pair of smart gloves that collected data from 156 sessions of the finger-tapping exam. To determine speed and amplitude in the finger tapping exam, we explored two different adaptive peak detection algorithms to detect peaks in the raw data. Human raters detected peaks to provide ground truth for analyzing the accuracy of the peak detectors. The average mean absolute error percentage for peak detection algorithms was 16%. We investigated the effect of medication wear off and our findings suggest a statistically significant difference in the dominant frequency and the peak-to-peak difference in time (p-value: 0.0256, 0.0366). A greater variability was also observed when medication effects were minimal. These findings are promising and encourage us to further investigate the smart gloves in a larger and longitudinal in-home study on Parkinson’s.