This study presents the development and validation of multifactor mathematical models that quantify how combined environmental stressors affect the reliability of fiber-optic communication components, including transceivers and passive modules. The model integrates four key input parameters — temperature (20–50 °C), relative humidity (60–90 %), Еlectromagnetic field (EM field) strength (1-5 V/m), and mechanical vibration (0.1-1.0 g) — to predict the main output quantities: optical power and reliability R(t). The regression and thermo-energetic formulations achieved a mean-squared-error of 0.024 W² and a correlation coefficient of R² = 0.91, while the reliability function estimated R(t) = 94 % after 12 months of average environmental loading. The deviation between modeled and simulated “measured” values did not exceed 1.6 % within the tested ranges, confirming the predictive adequacy of the approach. The models remain applicable for the environmental ranges listed above and can be integrated into real-time monitoring systems for adaptive reliability prediction of fiber-optic infrastructures.