The probability of failure of steel jacket platforms subjected to fatigue damage is computed by means of Monte Carlo simulations using limit state functions in which wave, wind, and deck loadings are expressed in terms of empirical functions of uncertain maximum wave height. Limit state functions associated with the base shear capacity of the jacket and the shear capacity of the deck legs were used. The sensitivity of the probability of failure to the coefficient of variation of resistance, of wave height, of resistance and loading biases, and to parameters in empirical loading functions, as well as the influence of the reserve strength ratio is analyzed using a simplified limit state function. Results from simulations are compared to those obtained with a formulation that relates the reserve strength ratio to the reliability index. An application to risk based inspection planning for extension of the service life of a platform is given.