Trends and variations in the frequency and strength of low-level jets can have a significant impact on the climatology of a region. Temporal analyses of low-level jets are complicated, however, by the many inhomogeneities in the rawinsonde record, particularly changes with time in the number of wind measurements in the vertical. The North American Regional Reanalysis (NARR), available from 1979-present, provides an alternative data source for climatological analyses of low-level jets. This study compares the frequency of low-level jets identified from vertical wind speed profiles at six rawinsonde stations in the central United States with the jet characteristics identified from wind profiles for the NARR grid point nearest the rawinsonde location. Four analysis years (1980, 1990, 2000, 2009) were chosen to capture the impact of changes with time in rawinsonde observational and archiving procedures on the jet characteristics. The six stations selected for the analysis had not been relocated during the study period. The NARR data set was found to substantially underestimate the frequency of low-level jets, by as much as 50 percent for some time periods and locations. The underestimation is particularly large in the later study years when rawinsonde observations have substantially greater vertical detail. Furthermore, the disparities in jet frequencies between the two data sets increase with the strength of the low-level wind maxima. Low-level jets were observed in both datasets on a relatively small number of observation times. Often, a low-level jet was observed in the rawinsonde wind profile but not in the NARR profile, although infrequently a jet was identified in the NARR, but not the rawinsonde, observations. Differences in jet frequency may in part be due to the relative strength of the wind in NARR and rawinsonde observations at different elevations. An analysis of paired NARR-rawinsonde observations for several stations and times showed that wind speeds in the lowest atmospheric levels were larger in NARR than in the rawinsonde observations, whereas at somewhat higher elevations rawinsonde wind speeds exceeded NARR speeds. The findings of this comparison suggest that climatological analyses of low-level jets based on NARR need to be interpreted cautiously.