Breakwater reliability is conventionally assessed one failure mode and typology at a time, leaving no common metric for ranking dissimilar coastal defence assets. This study introduces the Breakwater Reliability Index 2026 (BRI-2026), which aggregates Cornell reliability indices for three performance-based limit states (wave transmission exceedance, van der Meer armour stability and mean overtopping rate) through engineering-significance weights (0.35, 0.40, 0.25). A Gaussian-copula series-system failure probability and a four-class (A–D) rating anchored to target reliability values complement the index. As proof of concept, the framework is applied to twenty experimental and published records spanning five typologies: emerged, low-crested and submerged rubble mounds; impermeable trapezoidal barriers; floating units; plain and perforated vertical walls; and oscillating water column (OWC) breakwaters. Shapiro–Wilk testing rejects normality for five of seven hydraulic variables; AIC selects Weibull distributions for four variables and Gumbel for overtopping rate q, choices confirmed by bootstrap in 99.7% (q) and 93.0% (Kd) of resamples. Correlations confirm the energy-balance constraint (Kt–Kr, r = −0.870) and transmission–overtopping coupling (r = +0.693). With modal indices bounded at ±4.7, BRI-2026 spans −3.121 to +4.700; vertical walls and OWCs reach Class A or B, whereas submerged and low-crested mounds are Class D. A ±20% weight perturbation reclassifies at most two records, each within 0.17 of a threshold, and Monte Carlo treatment of weight and coefficient-of-variation uncertainty yields 90% intervals for all scores. With n = 20, results demonstrate the method rather than validate typology rankings; operational use requires a larger, source-documented database and independent field validation.