A differentially connected array of proton magnetometer operated within the epicentral region of the October 18, 1989, ML 7.1 Loma Prieta earthquake for 12 years from 1974 to 1986. The closest magnetometer stations was located 7.3 km from the epicenter of the earthquake and within 3 km of the site where anomalous ULF magnetic noise measurements were observed. Following the earthquake, the magnetometers were reinstalled with sensors replaced in the original undisturbed sensor holders. Comparison of pre-1986 total intensity magnetic field data with data obtained during the months following the earthquake indicate local offsets about 1 nT may have been generated at stations nearest the epicenter. Tests on other continuous differenced data frm 1983 to present indicate that offsets determined could be biased by as much 0.7 nT. The offsets can be approximately fit with a simple seismomagnetic model of the earthquake for which 1.9 m of right lateral and 1.3 m of dip slip (southwest side up) occurred on a fault patch between 6 km and 18 km deep and 45 km long. The total rock magnetization is assumed to be 1.5 A/m. Since the offset has persisted following the earthquake, an alternate explanation in terms of electokinetic effects is unlikely even though transient ground water flow occurred following the earthquake. Comparison of pre-1986 and similar post-seismic total magnetic field noise does not indicate any change caused by aliasing of ULF (0.01 Hz--10 Hz) magnetic noise in the vicinity of the Loma Prieta earthquake. ¿ American Geophysical Union 1990