Eighteen digital event recorders were deployed during May-June 1981 along the creeping-to-locked transition of the San Andreas fault zone near San Juan Bautista, California, as a supplement to the U.S. Geological Survey's central California seismic network. Analysis of 18 well-recorded microearthquakes (0.7 ≦ ML ≦ 2.8) located along the transition confirms the complexity of the crust and fault-zone structure of the transition region. Seismic-wave site amplification is 2 to 10 times greater at sites between the San Andreas and Sargent fault traces, consistent with other evidence for lower velocities in the upper 3 km of crust there. Routine mislocation of epicenters 2 to 4 km northeast of the Sargent fault are consistent with greater P-wave velocity northeast of the Sargent fault. Microearthquake source parameters are consistent with a more segmented and splayed fault geometry toward the northwest locked end of the transition. P-wave nodal planes for 10 microearthquakes located to the northwest, 9 on the Sargent fault, and 1 near the San Andreas, are oriented more westerly than nodal planes commonly obtained for the frequent moderate-size earthquakes on the creeping section of the San Andreas fault to the southeast. Static stress drop, dynamic stress drop, and apparent stress estimates all increase with seismic moment, with the apparent stress and dynamic stress drops equal to about 3 and 20 per cent, respectively of the static stress drop. Average fracture energies, calculated assuming complete stress drop, generally increase with source size (seismic moment, rupture area, seismic slip, etc.) from 7 to 3000 J/m2; the two events with anomalously low fracture energies occurred on the creeping section of the San Andreas fault.