The age-elevation pattern of thermochronologic data has been used for many years to estimate apparent erosion rates in orogenic settings. Recently, it has been suggested that thermochronologic data for detrital minerals from active river system sediments can be used as an effective proxy for bedrock age-elevation distributions, serving as an alternative method for erosion rate studies. This "detrital mineral thermochronology" (DMT) method is easier, faster, and more cost-effective than the traditional bedrock approach, and provides additional information about transients in topography and sediment delivery that may vary on diurnal to millennial timescales for a particular sampling site. However, two variants of the DMT method have been described in the literature, and they can yield very different erosion rate estimates. We tested both of these approaches against the traditional age-elevation method using detrital and bedrock 40Ar/39Ar muscovite data sets from a single river catchment in the Annapurna Range, Nepal Himalaya. A nominal erosion rate estimate of ~0.6 km/Myr for the 5.0--2.5 Ma period was calculated from the bedrock data, presented here for the first time. This result agrees with the ~0.7 km/Myr (maximum) estimate from the detrital data set, which was derived from the DMT variant that emphasizes the range of single-grain ages for a detrital sample, in this case, 11--2.5 Ma. However, the other DMT variant, which emphasizes the mean of the sample age distribution, yields an erosion rate estimate of ~2.3 km/Myr. The simplest explanation for this discrepancy is that erosion rate increased significantly after ~2.5 Ma, a scenario that is supported by apatite fission track data from the catchment.