Analysis of water levels in 21 wells in the Frenchman Flat area, Nevada Test Site, provides information on the accuracy of hydraulic-head calculations, temporal water-level trends, and potential causes of water-level fluctuations. Accurate hydraulic heads are particularly important in Frenchman Flat where the hydraulic gradients are relatively flat (less than 1 foot per mile) in the alluvial aquifer. Temporal water-level trends with magnitudes near or exceeding the regional hydraulic gradient may have a substantial effect on ground-water flow directions. Water-level measurements can be adjusted for the effects of barometric pressure, formation water density (from water-temperature measurements), borehole deviation, and land-surface altitude in selected wells in the Frenchman Flat area. Water levels in one well were adjusted for the effect of density; this adjustment was significantly greater (about 17 feet) than the adjustment of water levels for barometric pressure, borehole deviation, or land-surface altitude (less than about 4 feet). Water-level measurements from five wells exhibited trends that were statistically and hydrologically significant. Statistically significant water-level trends were observed for three wells completed in the alluvial aquifer (WW-5a, UE-5n, and PW-3), for one well completed in the carbonate aquifer (SM-23), and for one well completed in the quartzite confining unit (Army-6a). Potential causes of water-level fluctuations in wells in the Frenchman Flat area include changes in atmospheric conditions (precipitation and barometric pressure), Earth tides, seismic activity, past underground nuclear testing, and nearby pumping. Periodic water-level measurements in some wells completed in the carbonate aquifer indicate cyclic-type water-level fluctuations that generally correlate with longer term changes (more than 5 years) in precipitation. Ground-water pumping fromthe alluvial aquifer at well WW-5c and pumping and discharge from well RNM-2s appear to cause water-level fluctuations in nearby observation wells. The remaining known sources of water-level fluctuations do not appear to substantially affect water-level changes (seismic activity and underground nuclear testing) or do not affect changes over a period of more than 1 year (barometric pressure and Earth tides) in wells in the Frenchman Flat area.