The Picacho earth fissure transects subsiding alluvial sediments near the east periphery of the Picacho basin in south-central Arizona. The basin has undergone land subsidence of as much as 3.8 meters since the 1930's owing to compaction of the aquifer system in response to ground-water-Ievel declines that have exceeded 100 meters. The fissure, which extends generally north-south for 15 kilometers, exhibits horizontal tensile failure and as much as 0.6 meter of normal dip-slip movement at the land surface. The west side of the fissure is down thrown. The fissure was observed as early as 1927 and is the longest earth fissure in Arizona. Vertical and horizontal displacements were monitored along a line normal to the fissure. The survey line extends from a bedrock outcrop in the Picacho Mountains on the east, past an observation well near the fissure, to a point 1,422 meters to the west. From May 1980 to May 1984, the downthrown west side of the fissure subsided 167+-1.8 millimeters and moved 18+-1.5 millimeters westward into the basin. Concurrently, the relatively upthrown east side subsided 148+-1.8 millimeters and moved 14+-1.5 millimeters westward. Dislocation modeling of deformation along the survey line near the fissure indicates that dip-slip movement has occurred along a vertical fault surface that extends from the land surface to a depth of about 300 meters. Slip was 9 millimeters from May to December 1980 and also 9 millimeters from March to November 1981. Continuous measurements were made of horizontal movement across the fissure using a buried invar-wire horizontal extensometer, while water-level fluctuations were continuously monitored in four piezometers nested in two observation wells. The range of horizontal movement was 4.620 millimeters, and the range of water-level fluctuation in the nearest piezometer in the deep alluvium was 9.05 meters. The maximum annual opening of the fissure during the study period was 3.740 millimeters from March to October 1981, while the water level declined 7.59 meters. The fissure closed 1.033 millimeters from October 1981 to March 1982, while the water level recovered 6.94 meters. Opening and closing of the fissure were smooth and were correlated with water-level decline and recovery, respectively, recorded in the nearby piezometers. Pearson correlation coefficients between the water-level fluctuations in the deeper piezometers and horizontal movement ranged from 0.913 to 0.925. The correlogram with water-level decline as ordinate and horizontal strain as abscissa exhibits hysteresis loops for annual cycles of water-level fluctuation as well as near-vertical excursions for shorter cycles of pumping and recovery. Vertical and horizontal displacements also were monitored along a second survey line 1 kilometer north of and nearly parallel to the first survey line. The north line extends from bedrock on the east across three fissures to a point 582 meters to the west but does not cross the Picacho earth fissure. From May 1980 to May 1984, the fissure farthest from the mountain front along this line exhibited 20+-1 millimeters of opening and 33.3+-1.1 millimeters of vertical offset; the west side of the fissure was downthrown. During the same period, the zone between this fissure and the mountain front exhibited compression. The hypothesis of generalized differential compaction is supported by data taken at the study site for several reasons. First, the vertical offset across fissures and the fit of deformation to a dislocation model are consistent with an elastic model of differential vertical movement deep in the alluvium. Second, correlation is high between horizontal movement across the Picacho earth fissure and water-level fluctuations in the deeper local piezometers. Third, correlation is high between horizontal movement across the fissure and compaction farther west in the basin. The hypothesis of rotation of a rigid plate is not supported because (1) fissures sometime