Pseudotachylyte and other uncommon fault rocks form during paleoseismic slip. Such rocks afford frustratingly rare opportunities to study directly the products of earthquake slip and to understand earthquake processes better. I submit that layered ultracataclasites that surround and separate principal slip surfaces, which are common on many mature faults exhumed from seismogenic depth (e.g., Punchbowl fault), belong in this group. Most slip on mature faults occurs in large earthquakes, so many common fault rocks should record all or parts of the seismic cycle. E.g., the slip budgets of the N and S San Andreas fault (SAF) are well explained by large paleo-earthquakes (~2/3 of the SAF length), and the North Anatolian system apparently has "unzipped" three times in earthquake sequences. Similar conclusions can be drawn from Wasatch fault scarps and from seismic moment release rates of some subduction zones. Mature low-angle normal detachment faults (10-50 km slip) deliver brittle footwall fault rocks directly to the surface without overprinting unroofing events, so may provide a key paleoseismic rock record. Earthquakes on these are sparse (some say absent), and a few likely creep microseismically but the too-short seismic record does not exclude rare earthquakes. Several, however, have pseudotachylyte and/or related fault scarps that unambiguously record paleo-earthquakes. The WSD footwall rose from ~5 km depth and the top characteristically is 0.25-1 m of layered ultracataclasite with multiple slip surfaces, above 1-5 m of massive cataclasite. Two sites are key: In Nude Wash, a ~75 cm-thick body is made mainly of pseudotachylyte layers 1-2 cm thick. In the W fork of Nolina Wash, ~5 m of layered cataclasite and adjacent plutonic protolith record (1) episodic, sequential formation of dozens of layers from overlying rocks, (2) creation (interseismic creep?) and destruction (chaotic coseismic slip?) of low-T (non-plastic) foliations, (3) reworking of cohesive cataclasite and ultracataclasite as clasts, and (4) cataclasite injection. Thus, WSD slip was commonly (mainly?) paleoseismic, and principal slip surfaces within ultracataclasite were likely the most common products. Mature faults showing this combination probably should be considered paleoseismic unless strong evidence exists to the contrary.