The Copper River, located in southcentral Alaska, drains an area of more than 24,000 square miles. About 30 miles above its mouth, this large river enters Miles Lake, a proglacial lake formed by the retreat of Miles Glacier. Downstream from the outlet of Miles Lake, the Copper River flows past the face of Childs Glacier before it enters a large, broad, alluvial flood plain. The Copper River Highway traverses this flood plain and in 1996, 11 bridges were located along this section of the highway. These bridges cross parts or all of the Copper River and in recent years, some of these bridges have sustained serious damage due to the changing course of the Copper River. Although the annual mean discharge of the lower Copper River is 57,400 cubic feet per second, most of the flow occurs during the summer months from snowmelt, rainfall, and glacial melt. Approximately every six years, an outburst flood from Van Cleve Lake, a glacier-dammed lake formed by Miles Glacier, releases approximately 1 million acre-feet of water into the Copper River. At the peak outflow rate from Van Cleve Lake, the flow of the Copper River will increase an additional 140,000 and 190,000 cubic feet per second. Bedload sampling and continuous seismic reflection were used to show that Miles Lake traps virtually all the bedload being transported by the Copper River as it enters the lake from the north. The reservoir-like effect of Miles Lake results in the armoring of the channel of the Copper River downstream from Miles Lakes, past Childs Glacier, until it reaches the alluvial flood plain. At this point, bedload transport begins again. The lower Copper River transports 69 million tons per year of suspended sediment, approximately the same quantity as the Yukon River, which drains an area of more than 300,000 square miles. By correlating concurrent flows from a long-term streamflow- gaging station on the Copper River with a short-term streamflow-gaging station at the outlet of Miles Lake, long-term flow characteristics of the lower Copper River were synthesized. Historical discharge and cross-section data indicate that as late as 1970, most of the flow of the lower Copper River was through the first three bridges of the Copper River Highway as it begins to traverse the alluvial flood plain. In the mid 1980's, a percentage of the flow had shifted away from these three bridges and in 1995, only 51 percent of the flow of the Copper River passed through them. Eight different years of aerial photography of the lower Copper River were analyzed using Geographical Information System techniques. This analysis indicated that no major channel changes were caused by the 1964 earthquake. A flood in 1981 that had a recurrence interval of more than 100 years caused significant channel changes in the lower Copper River. A probability analysis of the lower Copper River indicated stable areas and the long-term locations of channels. By knowing the number of times a particular area has been occupied by water and the last year an area was occupied by water, areas of instability can be located. A Markov analysis of the lower Copper River indicated that the tendency of the flood plain is to remain in its current state. Large floods of the magnitude of the 1981 event are believed to be the cause of major changes in the lower Copper River.