This paper describes the results from a comparative wavelet analysis of a 0.1, 40-level Parallel Ocean Program model simulation of the North Atlantic with coincident in situ measurements of sea surface height (SSH) and temperature. The wavelet analysis is used to examine the realism of the surface's variability and its high-frequency signals (less than a year). Along the coast, it shows that the model's simulated fields of SSH are realistic with the correlations to tide gauge measurements on the order of 0.8. The wavelet spectra show that the model replicates the observations' frequency space. Comparisons of the model's temperatures to temperatures from NOAA buoys north and south of the Gulf Stream show that, while not replicating the location of mesoscale features all the time, the model's energy in the strong mesoscale regions compares favorably to data. Due to the TOPEX/POSEIDON's (T/P) sampling, which requires large areal averages and because of the model's spontaneous eddy field, the evaluation of the simulation in the open ocean is less conclusive. The model does show similarities to the T/P data at high latitudes where the sampling by the sensor is denser and also on the eastern side of the basin with its lower mesoscale activity. Spatially, there are similarities in the amplitude of the signals between the model and the observations in areas with a significant signal in a given spectral band.