Using LiDAR data collected on the levees along the Rio Grande River in New Mexico and Texas, an algorithm has been developed to automatically extract longitudinal elevation profiles. This algorithm consists of a series of filters, interpretation of geophysical properties, and digitized levee centerlines. The series of filters, in order of operation, include an alignment buffer filter, bare-earth filter, sampling filter and a maximum value filter. The result of the filter configuration is a 3-D polyline that models the levee crest. This algorithm allows for efficient identification of portions of levee that are lower than original design specifications. A comparison between the LiDAR levee crown extraction filter and a least-cost-path technique are offered.