We report the results of measurements of low frequency magnetic noise by two independent monitoring systems prior to the occurrence of the MS 7.1 Loma Prieta earthquake of 17 October 1989. Our measurements cover 25 narrow frequency bands in the more than six-decade frequency range 0.01 Hz–32 kHz, with a time resolution varying from a half hour in the ULF range (0.01–10 Hz) to one second in the ELF/VLF range (10 Hz–32 kHz). The ULF system is located near Corralitos, about 7 km from the epicenter. The ELF/VLF system is located on the Stanford campus, about 52 km from the epicenter. Analysis of the ELF/VLF data has revealed no precursor activity that we can identify at this time. However, the ULF data have some distinctive and anomalous features. First, a narrow-band signal appeared in the range 0.05–0.2. Hz around September 12 and persisted until the appearance of the second anomalous feature, which consisted of a substantial increase in the noise background starting on 5 October and covering almost the entire frequency range of the ULF system. Third, there was an anomalous dip in the noise background in the range 0.2–5 Hz, starting one day ahead of the earthquake. Finally, and perhaps most compelling, there was an increase to an exceptionally high level of activity in the range 0.01–0.5 Hz starting approximately three hours before the earthquake. There do not appear to have been any magnetic field fluctuations originating in the upper atmosphere that can account for this increase. Further, while our systems are sensitive to motion, seismic measurements indicate that there were no significant shocks preceding the quake. Thus, the various anomalous features in our data, and in particular the large-amplitude increase in activity starting three hours before the quake, may have been magnetic precursors.