Channel change and suspended-sediment transport were monitored in the Salt and Hassayampa Rivers in Maricopa County, Arizona, during the winter and summer rainy seasons of 1991-92. Flows were moderate. Results illustrate the high instability of these channels and high variability of process and response. A channelized, gravel-paved reach of the Salt River in an industrial part of Phoenix was incised 2 meters by sustained winter flows from upstream reservoir releases that had a peak discharge of 368 cubic meters per second. Similar amounts of channel incision occurred at bridge crossings at four other locations within 20 kilometers upstream from the study reach at 16th Street. Channel incision changed the stage-discharge relation at the streamflow-gaging station at 24th Street. Bank erosion below 16th Street undermined bank revetment and caused a large concrete-drop structure at the mouth of a storm drain to fall into the channel. About 23 kilometers upstream from the study area, bank erosion on the Salt River exhumed a landfill that resulted in entrainment and transport of refuse. The flows, which lasted 5 months beginning in early January, produced the highest peak discharge in 9 years on the normally dry lower Salt River. The flows were minor, however, compared to peak discharges that occurred during a series of floods from 1966 to 1980. The flood of 1980 that had a peak discharge of 5,100 cubic meters per second was the largest since 1905. In August 1992, several days of flows from reservoir releases produced a higher peak discharge of 493 cubic meters per second that resulted in little or no channel change. On a sandy, ephemeral reach of the Hassayampa River in rural Maricopa County west of Phoenix, as much as 20 meters of bank erosion resulted from three flows of short duration and low-to-moderate peak discharge. Most bank erosion resulted from a winter flow that lasted about 7 hours, had a peak discharge of 127 cubic meters per second, and an estimated recurrence interval of less than 5 years. A summer flow that lasted 3 hours had a peak discharge of 173 cubic meters per second and caused some bank erosion and possibly some dissection of terraces. The magnitude of change, however, was far less than that of the winter flow. Suspended-sediment concentration on the Salt River during the winter flows was typical of those for other regulated streams in Arizona and ranged from 2 to 617 milligrams per liter at discharges from 6.7 to 343 cubic meters per second. Fine-grained sediments in the channel bottom probably were the main source of sediment transported in suspension. During periods of prolonged, steady flows, suspended-sediment concentration tended to decline, which indicated a probable depletion of sediment supply. On the Hassayampa River, suspended-sediment concentrations ranged from 12,800 to 132,000 milligrams per liter at discharges of 13 to 128 cubic meters per second. The relation of sediment concentration to discharge was poor for the entire set of samples, but a clear pattern was evident for each period of storm runoff. In two of three periods of runoff sampled, maximum suspended-sediment concentration occurred just before peak discharge and declined rapidly.