Three years of continuous records of surface tilt preceding a moderate earthquake (ML = 4.3) on January 8, 1977, have been obtained at a point 5.5 km from the earthquake epicenter. A possible short-term percursive tilt to the southwest started December 18, 1976 and reached a maximum amplitude of 2 μradians relative to the tilt trend at this time. Other changes of this amplitude are evident, however, in the 3-year record. The sense of tilt changed abruptly following the earthquake, gradually returning to the general tilt trend. A substantial postseismic tilt of 10 μradians is consistent with aseismic slip of the Hayward fault, or on any of several other faults local to the tiltmeter in this region. The data are insufficient to discriminate between these possibilities, and accompanying surface displacements are apparently too small to be detected in the geodetic records. Short-term accelerated tilting just preceding the seismic events, as proposed by Wood and Allen (1971), are not apparent in these data. An observed coseismic tilt step of 0.14 μradians does not agree with that expected from current fault-failure models.