(Accepted for publication March 24, 1978) ABSTRACT: The U.S. Geological Survey has been operating a network of shallow-borehole tiltmeters in central California since June 1973. At six sites redundant instruments have been installed as a check on data coherency. These include the Sage Ranch, Tres Pinos, New Idria, Aromas, Bear Valley and San Juan Bautista tiltmeter sites. Preliminary results from the comparison of redundant data from the Aromas, Bear Valley and San Juan Bautista sites for periods of eight, three and seven months respectively, suggest that short-period tilt signals in the range 5 min T +8 to 10 +6 rad, but not including step offsets, show excellent agreement on closely spaced instruments. Agreement is not as good in this period range for instruments at San Juan Bautista with a separation of 200 m. Signals of interest observed in this period range include coseismic tilts, teleseisms and tilts associated with creep events. Tilt signals in the period range 3-5 h T +6 rad and wavelengths down to at least the instrument separation at the closely spaced sites (~several meters). Regarding longer-term coherency, the instruments at San Juan Bautista with 200-m spacing, agree within 0.5 μrad for the N—S component and 0.7 μrad for the E—W component for a period of two months. The closely spaced redundant instruments at Aromas agree within 2 μrad for the N—S component and 1 μrad for the E—W component for the eight-month period of operation. Data from the three sites have been checked for effects of temperature, atmospheric pressure and rainfall. The latter appears to be critically site dependent. The worst case tilts for 1 inch of rainfall can be more than 1 μrad with a duration of a few days to a week. Typical rain-induced tilts are less than 0.3 μrad for 1 inch of rain. The two instruments at the Sage Ranch site have been in operation for the longest period. However, they have shown local site or ground instability, high drift and lack of coherency since installation. Data are not yet available from the Tres Pinos or New Idria instruments. Deeper installation appears necessary for these instruments and two alternative methods of tiltmeter emplacement are currently being tested in an attempt to evaluate the depth, spatial and temporal dependency of surface tilt sources.