The WAM, the first proposed and most widely used third generation wave model, is employed to compare its estimated mean frequency against buoy measurements acquired in deep tropical ocean waters in Campos Basin, off Rio de Janeiro, Brazil. This model is noteworthy for being extensively tested and validated over a period of many years and is implemented operationally at several weather forecast centers in the world. However most of its validation tests were performed over a single wave parameter, that is significant wave height, using data covering all oceanic basins yielded by altimeters onboard satellites. Most, if not all of its assessment studies using detailed spectral buoy data are restricted to high latitude locations due to the sparseness of in situ measurements in low latitude waters. The standard frequency discretization used in the WAM cycle 4 model produces good scaling behavior for wind speeds ranging from 15 to 25 m/ s, which is way below the typical wind conditions encountered in tropical areas. It is shown that due to the standard frequency discretization employed the model underestimates the mean frequency of short waves measured in Campos Basin. This is indication that the cut-off frequency must be extended to a higher value, specially when the initial stages of wave growth are important.