Transpressional tectonics are typically associated with restraining bends on major active strike-slip faults, resulting in uplift and steep terrain. This produces dynamic erosional and depositional conditions and difficulties for established lines of palaeoseismological investigation. Consequently, in these areas data are lacking to determine tectonic behaviour and future hazard potential along these important fault segments. The Big Bend of the San Andreas fault in the Transverse Ranges of southern California exemplifies these problems. However, landslides, probably seismically triggered, are widespread in the rugged terrain of the Big Bend. Fluvial reworking of these deposits rapidly produces geomorphic planes and lines that are markers for subsequent fault slip. The most useful are offset and abandoned stream channels, for these are relatively high precision markers for identifying individual faulting events. Palaeoseismological studies from the central Big Bend, involving 14C ages of charcoal fragments from trench exposures, illustrate these points and indicate that the past three faulting events, including the great 1857 earthquake, were relatively similar in scale, each producing offsets of about 7–7.5 m. The mean recurrence interval is 140–220 years. The pre-1857 event here may be the 1812 event documented south of the Big Bend or an event which took place probably between 1630 and 1690. Ground breakage in both events extended south of the Big Bend, unlike the 1857 event where rupture was skewed to the north. The preceding faulting event ruptured both to the north and south of the Big Bend and probably occurred between 1465 and 1495. All these events centred on the Big Bend and may be typical for this fault segment, suggesting that models of uniform long-term slip rates may not be applicable to the south-central San Andreas. A slip-rate estimate of 34–51 mm a− 1 for the central Big Bend, although uncertain, may also imply higher slip in the Big Bend and highlights difficulties in correlating slip-rates between sites with different tectonic settings. Slip rates on the San Andreas may increase within the broad compressional tectonics zone of the Big Bend, compared to the north and south where the plate boundary is a relatively linear and sub-parallel series of dominantly strike-slip faults. Partitioning slip within the Big Bend is inherently uncertain and insufficient suitably comparable data are available to sustain a uniform slip model, although such models are a common working assumption.