Distribution of anglers along the banks of the Kenai River during 1997 was 60.5% on private land and 39.5% on public land. Land use by anglers in reach 1 (Kenai Lake to Skilak Lake) was 98.8% public. For reaches 2 (Skilak Lake to Moose River) and 4 (Soldotna Bridge to Warren Ames Bridge), use of public and private lands was nearly equal. For reach 3 (Moose River to the Soldotna Bridge), anglers tended to use private land more often, 60.4%. A comparison of public and private land use by anglers in 1996 and 1997 showed no change in reaches 1 and 2. A significant change was detected in reach 3 (2 = 59.9, P < 0.001) with a trend showing a 10% increase in angler use of public lands. There was a 7.6% decrease in use of public lands in reach 4 ( 2 = 23.3, P < 0.001). Of anglers observed, 86.9% fished from mainland banks. Of those anglers, 13.3% fished from boardwalks or docks, 58.2% stood in the water, and 28.5% fished while standing on the bank. No significant relationship was detected between angler traffic and bank integrity variables (bank angle, undercut bank, overhanging vegetation, and stream depth) at habitat survey sites. There was a significant difference detected between bank angle and habitat type (F = 8.22, P = 0.02). Prefishery and postfishery changes in bank angle showed a general decrease in bank angle for shrub/herbaceous sites and an increase in bank angle for herbaceous sites. Review of the data suggested that changes in bank angle and undercut bank tend to be a cyclic phenomenon associated with erosion cycles and that anglers may accelerate the process. Improved sampling may better define the relationship of angler impact with bank angle, undercut bank, and overhanging vegetation. Stream depth changes were a function of seasonal flow rather than angler impacts. The ability to measure erosion through the use of erosion pins was ineffective due to angler tampering. However, large areas of bank were documented to have calved at habitat survey sites receiving angler use. There was no significant change detected between angler traffic and vegetation sampled with departure from the riverbank, but there was a significant change detected for each habitat type for vegetation sampled within 5 feet of the bank (herbaceous: F = 4.12, P = 0.01; and shrub/herbaceous: F = 2.40, P = 0.01). The nearshore area showed a postfishery trend of decreasing vegetative cover. A better sample design for the transects departing from the bank would better define the relationship between angler impacts and vegetation changes. Penetrability measurements (soil resistance through the use of a penetrometer) were used as an indicator of soil compaction. There were no significant differences detected between angler traffic and penetrability measurements taken at 1 in, 3 in, and 6 in soil depths. Improved sample design may better define this relationship. Trampling was assessed by photo imagery analysis of prefishery and postfishery photographs of permanent vegetation plots. Of the cover classes assessed, there were no significant differences detected between angler traffic with percent cover for bare ground and water. There were significant differences detected between angler traffic with percent cover of vegetation (P < 0.01) and litter (P < 0.01). Further analyses of these two cover classes by habitat type detected a significant difference for herbaceous habitats for vegetative cover (P < 0.01) and litter cover (P < 0.01), but not for shrub/herbaceous habitats.