Background and Objectives: This study aimed to examine the responsiveness and concurrent validity of a serious game and its correlation between the use of serious games and upper limbs (UL) performance in Parkinson’s Disease (PD) patients. Materials and Methods: Twenty-four consecutive upper limbs (14 males, 8 females, age: 55–83 years) of PD patients were assessed. The clinical assessment included: the Box and Block test (BBT), Nine-Hole Peg test (9HPT), and sub-scores of the Unified Parkinson’s Disease Rating-Scale Motor section (UPDRS-M) to assess UL disability. Performance scores obtained in two different tests (Ex. A and Ex. B, respectively, the Trolley test and Mushrooms test) based on leap motion (LM) sensors were used to study the correlations with clinical scores. Results: The subjective fatigue experienced during LM tests was measured by the Borg Rating of Perceived Exertion (RPE, 0–10); the BBT and 9HPT showed the highest correlation coefficients with UPDRS-M scores (ICCs: −0.652 and 0.712, p < 0.05). Exercise A (Trolley test) correlated with UPDRS-M (ICC: 0.31, p < 0.05), but not with the 9HPT and BBT tests (ICCs: −0.447 and 0.390, p < 0.05), while Exercise B (Mushroom test) correlated with UPDRS-M (ICC: −0.40, p < 0.05), as did these last two tests (ICCs: −0.225 and 0.272, p < 0.05). The mean RPE during LM tests was 3.4 ± 3.2. The evaluation of upper limb performance is feasible and does not induce relevant fatigue. Conclusions: The analysis of the ICC supports the use of Test B to evaluate UL disability and performance in PD patients, while Test A is mostly correlated with disability. Specifically designed serious games on LM can serve as a method of impairment in the PD population.