Extreme loads of wind turbines are historically difficult to predict through simulation due to uncertainty in input conditions as well as in the simulation models. In addition, many long time series must be simulated for the statistics of the peak loads to become stationary. Offshore wind turbines require even more simulation due to the addition of stochastic wave loading. Floating offshore wind turbines, the subject of this paper, experience free-body motion as a result of wind and wave loading, and the phasing of wind turbulence, turbine motion, and large waves can be very influential in determining extreme loading. The International Electrotechnical Commission's 61400-3 standard covers loads analysis of offshore wind turbines, including only cursory references toward floating offshore wind turbines. This IEC design standard requires six 1-hour simulations to estimate extreme loads, which is not long enough for convergence of the statistics of peak loads for offshore turbines, especially floating turbines, which have higher and more variable loads due to platform motion. In this paper, 50-year wind and wave conditions are synthesized from data from the National Oceanographic and Atmospheric Administration's (NOAA's) floating data buoys for a suite of ocean sites suitable for floating or fixed bottom offshore wind turbines. The simulation software used in this paper is FAST, developed by the National Renewable Energy Laboratory, which is a coupled aero-hydro-servo-elastic wind turbine design tool. The current version of FAST which is used in this research includes second-order hydrodynamic effects, which may be important sources of loading in extreme conditions for certain floating platforms. TurbSim is used to create full-field turbulent wind files using the mean wind speed determined from the buoy data, while the hydrodynamics module within FAST handles the wave height time series using a JONSWAP spectrum. The OC3 spar buoy and the OC4 semi-submersible floating platforms are used as examples of realistic platform designs and a monopile is used as the fixed bottom example. A large number of 1-hour simulations are run to determine the convergence characteristics of each platform at each ocean site. These results are discussed and recommendations for future revisions of the design standard are made. Future work concerning various methods that will reduce the simulation cost of determining the converged extreme load will also be discussed.