Erosion of navigation canal banks is a direct cause of land loss, but there has been little quantitative analysis to determine why certain major canals exhibit faster widening rates (indicative of erosion) than others in the coastal zones of Texas, Louisiana, Mississippi, and Alabama. We hypothesize that navigation canals exhibit varying rates of erosion based on soil properties of the embankment substrate, vegetation type, geologic region (derived from digital versions of state geologic maps), and the presence or absence of canal bank armaments (that is, rock rip-rap, concrete bulkheads, or other shoreline protection structures). The first objective of this project was to map the shoreline position and substrate along both banks of the navigation canals, which were digitized from 3 different time periods of aerial photography spanning the years of 1978/79 to 2005/06. The second objective was to quantify the erosion rates of the navigation canals in the study area and to determine whether differences in erosion rates are related to embankment substrate, vegetation type, geologic region, or soil type. To measure changes in shoreline position over time, transects spaced at 50-m (164-ft) intervals were intersected with shorelines from all three time periods, and an annual rate of change was calculated for each transect. Mean annual rates of shoreline change ranged from 1.75 m/year (5.74 ft/year) on the west side of the Atchafalaya River, La., where there was shoreline advancement or canal narrowing, to -3.29 m/year (-10.79 ft/year) on the south side of the Theodore Ship Channel, Ala., where there was shoreline retreat or erosion. Statistical analysis indicated that there were significant differences in shoreline retreat rates according to geologic region and marsh vegetation type, and a weak relationship with soil organic content. This information can be used to better estimate future land loss rates associated with navigation canals and to prioritize the location of restoration and erosion mitigation efforts. Combining all canals together, our results also showed that canal erosion rates have slowed in recent years, with an average canal widening rate of -0.99 m/year (-3.25 ft/year) for the 1996/98-2005/06 time period compared to -1.71 m/year (-5.61 ft/year) for the earlier 1978/79-1996/98 time period. Future research could focus on obtaining detailed vessel traffic information for individual canals, which is likely a factor that influences canal bank erosion rates.