Eighteen years of seasonal 10-member ensemble simulations have been performed with observed sea surface temperature (SST) to assess the ability of dynamical models to predict seasonal-interannual climate variations during boreal summer. In addition, test cases have been designed to assess the role in climate trends, predictability of land initial conditions, systematic errors of precipitation and radiation fluxes at the land surface, and increasing CO₂. The model reproduces a global warming trend in surface temperature similar to that observed. This appears to be attributable to the influence of SST. However, surface flux replacement and atmospheric initialization can reduce the SST-driven trend even while improving the spatial pattern of surface air temperature. The results also show that land surface interaction has an effect both on surface temperature and the higher levels of the troposphere. However, realistically increasing CO₂ concentrations have little impact in these seasonal simulations.