The effectiveness of cognitive rehabilitation (CR) in Parkinson's disease (PD) is in its relative infancy, and nowadays there is insufficient information to support evidence-based clinical protocols. This study is aimed at testing a validated therapeutic strategy characterized by intensive computer-based attention-training program tailored to attention deficits. We further investigated the presence of synaptic plasticity by means of functional magnetic resonance imaging (fMRI). Using a randomized controlled study, we enrolled eight PD patients who underwent a CR program (Experimental group) and seven clinically/demographically-matched PD patients who underwent a placebo intervention (Control group). Brain activity was assessed using an 8-min resting state (RS) fMRI acquisition. Independent component analysis and statistical parametric mapping were used to assess the effect of CR on brain function. Significant effects were detected both at a phenotypic and at an intermediate phenotypic level. After CR, the Experimental group, in comparison with the Control group, showed a specific enhanced performance in cognitive performance as assessed by the SDMT and digit span forward. RS fMRI analysis for all networks revealed two significant groups (Experimental vs Control) × time (T0 vs T1) interaction effects on the analysis of the attention (superior parietal cortex) and central executive neural networks (dorsolateral prefrontal cortex). We demonstrated that intensive CR tailored for the impaired abilities impacts neural plasticity and improves some aspects of cognitive deficits of PD patients. The reported neurophysiological and behavioural effects corroborate the benefits of our therapeutic approach, which might have a reliable application in clinical management of cognitive deficits.