The overall objective of the current research is the development of a computationally efficient, conceptually simple, easy-to-use method providing loss estimates and other performance metrics for structural systems subjected to seismic loads. The method is based on (1) a novel probabilistic site-specific seismological model, (2) an efficient algorithm for calculating response statistics and (3) probabilistic models for life-cycle structural performance. The proposed seismic-hazard model uses earthquake records at the site of interest and the specific barrier seismological model to provide a more realistic representation of site seismic hazard. The information provided by records and the specific barrier model is aggregated in a Bayesian framework and used subsequently to simulate ground-motion samples as a function of the moment magnitude m and source-to-site distance r. Structural response statistics to simulated ground acceleration records are obtained by a novel efficient, nonintrusive method that resembles the Monte Carlo approach. Like Monte-Carlo, the method calculates structural responses to samples of the ground-motion process. Unlike Monte-Carlo, which uses a large number of samples selected at random, the proposed method uses a small number of samples selected in an optimal way. The efficiency of the proposed method allows calculation of distributions, rather than just mean values, for downtime cost, damage, and other metrics. Probability distributions can be used in insurance applications to calculate premiums to cover cost of damage. They are also essential tools for assessing tail risk, a quantity which accounts for low-probability events with high impact, used for transferring risk to reinsurance markets. These capabilities are particularly important when dealing with extreme events. Numerical results are presented for linear and non-linear systems. Life-cycle scenarios for seismic events are simulated and used to estimate life-cycle cost and damage.