"There are a few empirical models that predict wave parameters in shallow waters. This report compares two empirical models : Shore Protection Manual model (SPM, 1984) modified later by Hurdle and Strive (1989); and Krylov model (described in Massel, 1996) with field data recorded with Datawell waverider buoy, installed at the depth of 22 m LAT, on the edge of Keppel Bay (Central Queensland, Australia). Four years of recorded data was analysed. Approximately 430 wave conditions covering the range of significant wave data from 0.5 m to 3.5 m; and peak period from 3 to 13 s, with duration of over 9 hrs in each case was selected for detailed analysis. Not far from the wave-recording buoy is a weather station located on Rundle Island where wind conditions have been recorded automatically every 3 hours. Analysis has shown that both models significantly overestimate predicted wave heights and periods compared with recorded ones. Vincent and Hughes (1985) formula for significant wave height in the depth-control wave train, when modified by the author by including dispersion of waves, represents recorded data very accurately. In addition waves with long periods (>9 sec), considered as swell waves were analysed to evaluate predominant swell direction. The proper knowledge of dominant swell waves' direction is important in consideration of coastline realignment as a sustainable "soft" method of beach stabilization."--p. i.