Airborne bathymetric LiDAR surveys generally cover areas somewhat larger than the exact area of interest to ensure complete coverage. The result is the collection of some LiDAR data that do not contain extractable bathymetric soundings (due primarily to water depth).  Moreover, that such data do not contain bathymetric soundings currently can only be determined by processing each LiDAR tile viewing the a posteriori results. This work describes a method for a priori identification of 500 m-by- 500m LiDAR data tiles unlikely to have extractable bathymetry thereby eliminating the need to process some tiles. The approach quantitatively analyses the frequency distribution of the unprocessed sounding depths on a tile to estimate the likelihood that it contains extractable bathymetry.  For a test data set of 1260 tiles, the prediction was 91% accurate with 5% false negatives and 4% false positives. Operationally this could translate to a 25% (360 tiles) reduction in the number of tiles requiring bathymetric processing. This method will be presented and explained, its potential operational usage discussed, and implications of false negatives and false positives explored.