This paper documents the first systematic assessment for one aspect of paleoflood analysis: the relation between paleostage indicators (PSIs) and the peak stage of floods responsible for their emplacement. A main source of uncertainty in paleoflood reconstructions is maximum flood stage inferred from PSIs. Typically, the elevation of the top of the PSI is used as the minimum elevation of the flood that deposited the sediments, but little evidence supports this assumption. Data from recent large floods (median recurrence interval of 75 years), primarily in the western and west-central United States, were used for a comprehensive evaluation of the relation between flood-transported sediments (fresh PSIs) deposited as flood bars and/or slackwater deposits and flood high-water marks (HWMs). Surveys of flood-deposited sediments, HWMs, and channel geometry were made for 192 sites, primarily streams with gradients larger than 0.002 m/m and depths less than about 4.5 m. No statistically significant relation was identified with channel gradient, particle size, discharge, recurrence interval, expansion or contraction ratios, width, or depth variables. Analysis of the data indicates that the elevation at the top of the flood sediments (PSIs) is on average 15 mm higher than the HWMs; results slightly vary with type of deposit. The elevation at the top of flood-deposited sediments (new PSIs), preferably deposits nearest to channel margins, provides a reliable and accurate indication of the maximum height of the flood. Forty-six peak discharges that we estimated with the critical-depth and slope-conveyance methods also were compared with peak discharges independently made at streamflow-gaging stations, primarily determined using current-meter measurement; forty-two of the 46 estimates were within ±25 percent of the gage peak discharge measurements.