The differential diagnosis of Parkinson's diseases (PD) is challenging, especially in the early stages of the disease. We developed a microRNA profiling strategy for exosomal miRNAs isolated from cerebrospinal fluid (CSF) in PD and AD. Sixteen exosomal miRNAs were up regulated and 11 miRNAs were under regulated significantly in PD CSF when compared with those in healthy controls (relative fold > 2, p < 0.05). MiR-1 and miR-19b-3p were validated and significantly reduced in independent samples. While miR-153, miR-409-3p, miR-10a-5p, and let-7g-3p were significantly over expressed in PD CSF exosome. Bioinformatic analysis by DIANA-mirPath demonstrated that Neurotrophin signaling, mTOR signaling, Ubiquitin mediated proteolysis, Dopaminergic synapse, and Glutamatergic synapse were the most prominent pathways enriched in quantiles with PD miRNA patterns. Messenger RNA (mRNA) transcripts [amyloid precursor protein (APP), α-synuclein (α-syn), Tau, neurofilament light gene (NF-L), DJ-1/PARK7, Fractalkine and Neurosin] and long non-coding RNAs (RP11-462G22.1 and PCA3) were differentially expressed in CSF exosomes in PD and AD patients. These data demonstrated that CSF exosomal RNA molecules are reliable biomarkers with fair robustness in regard to specificity and sensitivity in differentiating PD from healthy and diseased (AD) controls.