There exist many different human cancers, but regardless of the cancer type, an early diagnosis is a necessary condition for further optimal outcomes from the disease. Therefore, efficient specific and sensitive cancer biomarkers are urgently needed. This is especially true for the cancers depicting a silent progression, and those only diagnosed in an already metastatic state with a poor survival prognostic. After a rapid overview of the previous methods for cancer diagnosis, the outstanding characteristics of extracellular vesicles (EVs) will be presented, as new interesting candidates for early cancer diagnosis in human biofluid non-invasive liquid biopsy. The present review aims to give the state-of-the-art of the numerous searches of efficient EV-mediated cancer diagnosis. The corresponding literature quest was performed by means of an original approach, using a powerful Expernova Questel big data platform, which was specifically adapted for a literature search on EVs. The chosen collected scientific papers are presented in two parts, the first one drawing up a picture of the current general status of EV-mediated cancer diagnosis and the second one showing recent applications of such EV-mediated diagnosis for six important human-specific cancers, i.e., lung, breast, prostate, colorectal, ovary and pancreatic cancers. However, the promising perspective of finally succeeding in the worldwide quest for the much-needed early cancer diagnosis has to be moderated by the many remaining challenges left to solve before achieving the efficient clinical translation of the constantly increasing scientific knowledge.