Prior to emergence as fry, salmonid embryos incubating within gravel nests called “redds” are vulnerable to substrate mobilization and lowering of the streambed, a process termed “streambed scour,” during floods. Water managers regulating discharge in salmonid-bearing rivers need information about the magnitude of discharge during which the scour of substrate surrounding salmonid redds occurs. The time when scour occurs, however, is difficult to measure and usually poorly constrained. The South Fork Tolt River in western Washington supplies the City of Seattle with hydroelectric power and about 40 percent of its municipal water needs, while providing spawning habitat for two salmonid species listed under the Endangered Species Act: Chinook salmon (Oncorhynchus tshawytscha) and steelhead trout (O. mykiss). The U.S. Geological Survey, in cooperation with Seattle City Light and Seattle Public Utilities, began a study in 2015 using accelerometer scour monitors (ASM) to characterize the timing of and hydrologic conditions associated with streambed scour at the depth of incubating salmonid embryos in the South Fork Tolt River. Prior to this study, operational thresholds for peak discharge on the South Fork Tolt River were 350 cubic feet per second (cfs) in the upper part of the river and 550 cfs in the lower part of the river as measured at USGS streamgages 12148000 and 12148300, respectively. These thresholds were developed from the peak discharge associated with observations of the flattening of redd structure and not from direct measurement of scour at the depth of egg pockets within redds. Accelerometer scour monitors were deployed at the level of salmonid egg pockets in spawning habitat of the South Fork Tolt River to record the temporal pattern of streambed scour at the depth of incubating salmon eggs during fall and winter flood seasons of water years (WY) 2016 and 2017. Thirteen of 48 ASMs deployed during the WY 2016 flood season recorded scour attributed to high streamflow when discharge measured at USGS streamgage 12148300 (the lower river streamgage used as an index gage) was between 969 and 1,360 cfs. Local discharge at individual scour sites varied depending on the timing of tributary inputs and downstream transport of water. During the subsequent flood season in WY 2017, peak discharge at the index gage reached 809 cfs. None of the 38 ASMs deployed recorded scour attributed to streamflow alone, although 10 ASMs recorded localized bed movement attributed to spawning activity of fish. Most scour at the depth of redds measured during WY 2016 occurred at or before peak flood discharge consistent with previous redd scour studies. The lack of scour measured in WY 2017 when peak discharge (809 cfs) was less than the minimum discharge when scour occurred in WY 2016 (969 cfs) suggests minimal to no scour of egg pockets in salmonid redds when discharge is less than 809 cfs.