BACKGROUND: Parkinson's disease (PD) encompasses motor (e.g., bradykinesia) and non-motor (e.g., apathy) symptoms. OBJECTIVE: We aimed to use reflexive and voluntary saccades as a proxy for bradykinesia and apathy. METHODS: Seventeen PD patients and thirteen controls (matched for age and educational level) were recruited. We assessed apathy using the Dimensional Apathy Scale (DAS) and bradykinesia using MDS-UPDRS III. Subjects were asked to fixate successively two green points (cues, 40° apart) alternating at 1 Hz. After 20 s, all stimuli disappeared, and participants were required to continue fixating on the previous locations of the cues at the same frequency for another 20 s. We measured the Maximal Amplitude (MA) (saccade amplitude from side to side) and its period. Linear mixed models assessed the effect of the group (patient/control), cue, DAS, and bradykinesia score. RESULTS: Overall, the DAS was similarly correlated to the period (p = 0.0157) and the MA (p = 0.0002) in the absence of a cue. However, this correlation was significant only in the patient subgroup for the MA (p = 0.0005). In the absence of cue, bradykinesia was similarly correlated to the period (p = .0002) and the MA (p = 0.0004). However, the period was better correlated to bradykinesia than the DAS. CONCLUSIONS: While the saccade period best correlates with bradykinesia, maximal amplitude in the absence of cue better reflects the severity of apathy. Our paradigm may be a promising objective biomarker for assessing bradykinesia and apathy in PD.