The central Taiwan thrust belt is an important laboratory for study of fault and crustal mechanics because its structure is well constrained and it is the site of the well-instrumented 1999 Chi-Chi earthquake (Mw 7.6) and the post-Chi-Chi TCDP scientific drill holes. Here we summarize recent results on crustal and fault strength within this active thrust belt, based on critical-taper wedge mechanics and published TCDP temperature anomalies and stress measurements. These constraints indicate that the major faults are exceedingly weak with effective friction of approximately 0.03-0.1, whereas the deforming crust containing them is strong (sigma_{1}-sigma_{3})/ rho gz approximately 0.6-1. Thus there is an order of magnitude difference between fault strength and crustal strength sigma_{tau}/(sigma_{1}-sigma_{3}) approximately 0.1. Furthermore, the extreme fault weakness is not caused by regional ambient high pore-fluid pressures as classically proposed by the Hubbert and Rubey because petroleum bore-hole data show that the fluid pressures are regionally hydrostatic. These results underline the outstanding causal questions of fault weakness and crustal strength.

The classic graph of crustal strength as a function of depth shows linearly increasing brittle strength above the brittle-plastic transition. This linear increase is a consequence not only of the pressure dependence of brittle strength but also an assumption that the depth-normalized pore-fluid pressure lambda = P_{f}/ rho_{r} gz is constant, which is perhaps only plausible in the case of hydrostatic pore-fluid pressures (lambda approximately 0.4). Much deep bore-hole stress data agrees with this assumption, showing the predicted linear increase in strength together with stress magnitudes that are consistent with hydrostatic pore-fluid pressures. In contrast, observed pore-fluid pressures in deeper parts of deforming clastic sedimentary basins and active plate-boundary mountain belts are commonly in excess of hydrostatic. These deforming sedimentary basins typically have pore-fluid pressures that are dominated by disequilibrium compaction, showing fully compacted sediments with hydrostatic fluid pressures at shallow depths until the fluid-retention depth z_{FRD} is reached, below which sediments are increasingly undercompacted and overpressured. For this disequilibrium-compaction mechanism, the fractional brittle weakening (1-lambda) below the fluid-retention depth is a simple function of depth (1-lambda) approximately 0.6(z_{FRD}/z), which directly leads to a predicted crustal-strength profile that is radically different from the classic hydrostatic profile. The brittle strength below the fluid-retention depth is predicted to be constant and equal to the strength at the fluid-retention depth. Some stress measurements in deeper parts of sedimentary basins appear consistent with this constant-strength prediction. Furthermore, observations from western Taiwan show that $z_{FRD}$ is fixed relative to the land surface during active uplift and erosion, therefore crustal strength should be approximately unchanged by exhumation except for cohesive effects. The full limits of the disequilibrium-compaction regime are not well know. However it is only as non-hydrostatic pore-fluid pressures decay that deformed sedimentary mountain belts are expected to show crustal strengths similar to the classic graph.