Due to extracellular vesicles role as intracellular messengers and potential as diagnostic tools, exosome research has gained much attention over the last decade. EV (Extracellular vesicle) enrichment in clinical cohorts needs to be investigated on a large scale, but the lack of rapid, reproducible, efficient, and low-cost methods is a major obstacle. The advancement in microfluidics has provided an excellent opportunity for shifting from conventional sub-micronsized isolation and purification methods to more robust and cost-effective lab-on-chip platforms. The acoustic-driven separation approach applies differential forces acting on target particles, guiding them towards different paths in a label-free and biocompatible manner. The objective of this thesis is developing an acoustic-based microfluidic platform for the separation of different subgroups of the extracellular vesicles under a label-free and contact-free manner. In the first part of the study an acoustofluidic separation platform was developed and optimized in terms of electrical and mechanical characteristics for the nanoscale particle manipulation. It was concluded that the return loss and acoustic window are two main critical factors influencing the radiation force and ultimately particle’s trajectory in the microchannel. We demonstrated that by utilizing the machine learning technique the design of the IDT (interdigital transducer) features of the acoustic resonator can be automated for achieving the highest separation performance. Next, we find that the concurrent application of dielectrophoretic (DEP) and acoustophoretic forces decreases the minimum particle separation size and eliminates the limitations associated with bubble generation and particle aggregation of the devices that solely rely on sound waves. This approach is then used to sort subpopulations of extracellular vesicles. In the end we demonstrate a sheath-less EVs separation device based on highly localized acoustic streaming actuation.