Extracellular vesicles (EVs) have emerged as a potential biomarker in liquid biopsy. However, cancer heterogeneity poses significant challenge to precise molecular diagnosis based on single-parameter input. Hence, strategies for analyzing multiple inputs with molecular computing were developed with the aim of improving diagnostic accuracy in liquid biopsy. In the present study, based on the surface of aptamer-encoded EVs, three toe-hold extended DNA aptamers served as specific inputs to perform AND-logic-gating to distinguish between healthy and cancerous EVs. In addition, this strategy has been successfully employed to analyze circulating EVs in clinical samples from colorectal cancer patients and healthy donors. The developed method has a promising future in the analysis of multiplex EV membrane proteins and the identification of early cancer.