Periodic inspection policy is performed for some degradation systems to check their degradation states, whereas it is usually difficult to implement on time due to impact of some random factors. Inspections and some maintenance actions are implemented in an inspection window with random, and thus how to optimize the inspection windows and the degradation threshold of the system to perform preventive maintenance (PM) are beneficial in practice. To this end, an optimisation of quasiperiodic inspection and PM policy with inspection window is proposed for a degradation system whose degradation followed Wiener process with a linear drift. Assume that PM can change the degradation rate and inspections are randomly performed in each inspection window. After optimisation, the optimal interval of the inspection window, the degradation threshold of PM and PM policy are determined by minimising the long-term running cost rate of the system. Finally, modeling and optimisation are illustrated using the degradation process of an axial piston pump, and the sensitivity analysis of some key parameters is conducted.