Since 1972, a number of static strain changes have been measured at Pi¿on Flat Observatory, southern California, as a result of regional faulting. As recorded by a collection of long-baselength strainmeters and til-meters, these signals are in good quantitative agreement with the deformation calculated for a dislocation in an elastic half-space. In contrast, colocated short-baselength sensors are found to produce spurious results whose size is much larger than expected theoretically and well correlated with peak site acceleration. Calculations show that the root-mean-square (rms) of the coseismic signals from several different instruments at one site, or from several sites at known hypocentral distances, provide a good estimate of the seismic moment, a result confirmed by the observations. Static moment estimates for the four largest events in this period are Imperial Valley 1979, M0=9.0¿1018 N m; Mexicali Valley 1980, M0=6.1¿1018 N m; Westmorland 1981, M0=5.3¿1017 N m; and Brawley 1979, M0=2.3¿1017 N m. These estimated moments generally exceed those determined seismically by about 25%, which is as expected since they include some postseismic deformation. Results from the Mexicali Valley earthquake are particularly interesting, indicating an average fault strike approximately 9¿ different from the historic fault trace. Several small to moderate earthquakes (ML 3.7--5.2) were also well enough recorded to estimate source parameters. For the only other detectable events in the period 1972--1982 (a total of three others for which the strain change Δϵrms was greater than 0.4 nϵ), high accelerations caused most of the instruments at the site to malfunction. No precursory signals were evident for any of the earthquakes recorded at this site. ¿ American Geophysical Union 1988