<p>The primary objectives of maintenance strategies revolve around cost reduction and reliability enhancement. This paper introduces an innovative condition-based maintenance (CBM) model specifically tailored for series–parallel systems characterized by interdependent component failures. In particular, the proposed model addresses multiple subsystems, each consisting of parallel components where operation depends on the functionality of at least k components arranged in series. To effectively capture the impact of stochastic covariate values on component degradation, a proportional hazards model (PHM) is employed. Additionally, a tampered failure rate (TFR) model is introduced to assess the reliability of load-sharing systems, featuring identical components with stochastic failure dependency. To model the complex relationships among component failures, transition probability matrices are utilized. This approach enables the determination of optimal control limits for each subsystem. The system undergoes periodic inspections, and based on these inspections, an optimal replacement policy is devised to minimize the total expected cost over the planning horizon. Overall, the integration of these elements into a cohesive model, considering the stochastic dependence among components, represents a level of computational and modeling complexity unparalleled in previous literature. Numerical examples, solved using MATLAB software, illustrate the practical applicability of the proposed model. The findings demonstrate a notable improvement in reliability, achieving approximately 99% compared to 91–95% in the base model, along with a 50–65% reduction in system costs. This underscores the effectiveness of integrating failure dependencies into maintenance strategies, providing enhanced reliability and cost-efficiency over the base model.</p>