A 2-yr set of profile data from Torrey Pine Beach, California, measured at monthly intervals has been statistically analyzed by using empirical eigenfunctions. The analysis separates the temporal and spatial dependence of the data, this separation permitting beach changes to be described objectively by a linear combination of corresponding time and space functions. Most of the variation in profile configuration can be accounted for by three eigenfunctions corresponding to the three largest eigenvalues. The largest eigenvalue corresponds to an eigenfunction called the 'mean beach function', which represents an average profile. A second eigenfunction, the 'bar-berm function', has a large maximum at the location of the summer berm and a minimum at the location of the winter bar, indicating its relation to the seasonal onshore-offshore movement of sand. The third eigenfunction, the 'terrace function', has a maximum at the location of the low-tide terrace. Results if this study indicate that the eigenfunctions are useful in the analysis of beach profile data and provide objective insight into the nature of the variations of the profile configuration.