The study presents a reliability-based fatigue life prediction model for the ocean current turbine rotor blades. The numerically simulated bending moment ranges based on the measured current velocities off the Southeast coast line of Florida over a one month period are used to reflect the short-term distribution of the bending moment ranges for an idealized marine current turbine rotor blade. The 2-parameter Weibull distribution is used to fit the short-term distribution and then used to obtain the long-term distribution over the design life. The long-term distribution is then used to determine the number of cycles for any given bending moment range. The published laboratory test data in the form of an ϵ-N curve is used in conjunction with the long-term distribution of the bending moment ranges in the prediction of the fatigue failure of the rotor blade using Miner's rule. The first-order reliability method is used in order to determine the reliability index for a given section modulus over a given design life. The results of reliability analysis are then used to calibrate the partial safety factors for load and resistance.