This paper focuses on series systems' dynamic opportunistic maintenance scheduling. Based on the machine-level predictive maintenance (PdM) method, a novel TOC–VLLTW methodology combined theory of constraints (TOC) policy and variable lead-lag time window (VLLTW) policy is proposed. The TOC policy provides machines' priorities according to their PdM durations to decrease system downtime when scheduling opportunistic maintenance. The VLLTW policy provides variable lead-lag time windows against different machines, allowing for more flexible and economic system opportunistic maintenance schedules. This proposed methodology is demonstrated through the case study based on the collected reliability information from a quayside container system. The results can effectively prove the effectiveness of the TOC–VLLTW methodology.