For the past 20 years, much effort has been directed to determining the present-day relative motion of the Pacific and North American plates using two independent approaches. One uses geologic observations and geodetic measurements along the San Andreas Fault and other faults in the plate boundary zone. The other is based on plate motion models that incorporate spreading rates from marine magnetic anomalies, transform azimuths, and earthquake slip vectors. Geologic and geodetic studies find two principal elements of deformation: slip along the San Andreas of ~34 mm/yr directed N41 ¿W, and extension across the Basin and Range province of about 10 mm/yr directed N56 ¿W. In contrast, plate motion studies find 56--60 mm/yr directed N35 ¿W. The discrepancy between these estimates, a vector of about 15 mm/yr orientated nearly due north, is often attributed to a combination of slip along faults parallel to the San Andreas and shortening normal to it.

Here we revise the estimate of Pacific-North America motion by analyzing marine magnetic profiles from the Gulf of California. Since 3 Ma, spreading has averaged 48 mm/yr, 10 mm/yr slower than estimated before, and consistent with the 49 mm/yr spreading predicted by a new global plate motion model derived without any data along the Pacific-North America boundary. The discrepancy with geodetic and geologic estimates is thus reduced to only 5 mm/yr parallel to the San Andreas, 60% less than estimated before, and 7 mm/yr of shortening across the San Andreas, similar to prior estimates. These results suggest that strike-slip motion on faults west of the San Andreas is less than thought before, a conclusion consistent with geodetic, siesmological, and other geologic observations. ¿ 1987 American Geophysical Union