In this paper, suffering from multiple unknowns including unmodeled dynamics, uncertainties, and disturbances, a fuzzy unknown observer-based robust adaptive path following control (FUO-RAPFC) scheme for an underactuated surface vehicle (USV) is proposed. Main contributions are as follows: (1) A surge-guided line-of-sight (SGLOS) guidance law is created to guide surge speed and heading angle, simultaneously, and thereby adapting to path following errors to enhance robustness and flexibility; (2) By virtue of combining auxiliary observation dynamics with adaptive approximation error compensation, the FUO is innovatively devised to accurately lumped unknowns rather than bounded observations; (3) FUO-based robust adaptive tracking control laws are eventually synthesized and ensure that the guided signals can be globally asymptotically tracked. Simulation studies are conducted to demonstrate remarkable performance of the FUO-RAPFC scheme.