The interaction between the unsteady cavitating flow and hydrodynamic performance around a pitching Clark-Y hydrofoil is investigated experimentally and numerically. The experiments were conducted in the looped cavitation tunnel, and the cavitation patterns are documented by two high-speed digital cameras, and the moment of hydrofoil is measured by the moment sensor. The pitching hydrofoil is controlled from alpha(+) = 10 degrees to alpha(+) = 15 degrees firstly, and goes back to alpha(-) = 5 degrees from alpha(+) = 15 degrees, then goes back to alpha(+) = 10 degrees from alpha(-) = 5 degrees at Re = 4.4*10(5). The upstream velocity U-infinity and the cavitation number sigma is fixed at 6.3 m/s and 1.38, respectively. And the pitching rate is (alpha) over dot = 40 degrees/s, (alpha) over dot* = 0:086. The numerical investigations were performed by solving the incompressible UNRANS equations via the commercial code CFX using the Merkle cavitation model, the coupled k-omega SST turbulence model and gamma-Re-theta transition model. The predicted cavity patterns and moment coefficients agree well with the experimental results. The results showed there are two distinct cavitation patterns (Multi-scale cloud cavitation and Traveling sheet cavitation). For the Multi-scale cloud cavitation phase (alpha(+) = 10 degrees-alpha(-) = 10 degrees), the re-entrant flow is the main factor on the stability of cavitating flow structures, which is responsible for different shedding patterns. According to the breaking position and re-entrant flow thickness, this stage is divided into three different patterns of the cavity development and shedding. For the Traveling sheet cavitation phase (alpha(-) = 10 degrees-alpha(+) = 10 degrees), the shedding of cavity mainly results from the interaction of the re-entrant flow and the fluctuation of the gas liquid interface, thus leading to the irregular breaking points. The cavitating flow structure of different phases were further investigated using the LESs, which will be help to identify the dynamic behavior of flow structures effectively. (C) 2020 Elsevier Ltd. All rights reserved.