The U.S. Coast and Geodetic Survey ran first-order level surveys in 1934, 1955, and 1967 across the fault zone associated with the 1954 sequence of earthquakes at Rainbow Mountain (M = 6.6 and 6.8), Dixie Valley (M = 6.8), and Fairview Peak (M = 7.1) in Nevada. The difference between the 1955 and 1967 surveys clearly shows distinct anomalies over distances of several kilometers at all but one of the fault scarps mapped after the 1954 earthquakes. The anomalies resemble deformation produced by normal faulting extending to a depth of at least several kilometers and, consequently, are interpreted as implying continued slip on the earthquake faults in a period beginning at least 6 months after the earthquake. At the Fairview Peak Fault, the inferred postearthquake slip is about 5 per cent of the displacement observed at the time of the earthquake. The difference between the 1955 and 1967 surveys suggests an overall tilt of 2 mm/km down to the west extending over a 90-km distance crossing the fault zone. The difference between the 1934 and 1955 surveys suggests an overall tilt of 0.8 mm/km down to the east extending over a 200-km section. However, these regional tilts might be due to unusually large systematic errors in the level surveys. Gravity and seismic-refraction surveys indicate that the region as a whole is isostatically compensated, although the mountain ranges and intervening basins are not individually compensated. Thus, the regional elevation changes, if they exist, cannot be accounted for by isostasy.