<p>This study proposes an integrated Bayesian-reliability framework to optimize irrigation and fertigation management in greenhouse strawberry cultivation, a system highly dependent on manual operations. The framework combines the Bayesian Best-Worst Method (BBWM) with reliability engineering to quantify performance and uncertainty in water- and fertilizer-sensitive processes. A Bayesian hierarchical model aggregates judgments from seven experts to determine critical weights for 20 performance indicators, which are then integrated into a Reliability Block Diagram (RBD) to model system interdependencies and compute overall system reliability. The results identify water and fertilizer management as the most critical single indicator (weight 0.0765), while leaf picking and flower/fruit thinning collectively account for over 57% of the total importance. A case study on a cooperating farm shows that the proposed approach quantifies the baseline system reliability as 0.783. Sensitivity analysis further reveals that temperature control during the fruiting stage is a key bottleneck affecting system reliability. Targeted improvements in irrigation scheduling and fertigation practices can enhance system reliability and significantly improve water use efficiency while reducing operational risks. This work provides a generalizable methodology for performance evaluation, bottleneck diagnosis, and resource optimization in controlled-environment agriculture, with direct implications for sustainable irrigation management.</p>