Within three days of the Coyote Lake earthquake, M = 5.7, of August 6, 1979, two recording magnetometers were installed on the Calaveras fault, 8 and 13 km south of the epicenter, in order to search for tectonomagnetic effects associated with aftershocks. These instruments supplemented the U.S. Geological Survey (USGS) permanent station and a temporary USGS station installed at about the same time. In addition, eight precise magnetic surveys were made during the month following installation at 30 sites in the region. After removal of ionospheric and magnetospheric disturbances from the records of the continuously recording magnetometers, no coseismic changes were evident greater than the noise level, which for hourly averages was less than 2.0 nanoteslas (nT) peak to peak at the time of the earthquakes. The majority of the differences in magnetic field between the survey sites and one of the continuously recording sites remained constant to ±2 nT. Differences at five sites which lie on the fault changed by slightly more than 4 nT. However, lack of coherence between the observed changes and the background seismicity casts doubt on interpreting them as tectonomagnetic effects.