The fault rupture nucleation point of the Irpinia earthquake is relocated, following the recent identification of —20 km of surface faulting, the Carpineta and Picentini fault scarps, in addition to the —15 km previously documented on the Marzano and San Gregorio faults, all of which have northwestward strike and —60° northeastward dip. This relocation, relative to a well-located aftershock, is based on detailed analysis of Pn-wave arrival times at permanent seismograph stations. It indicates a range of revised origin times and nucleation point positions at (10±12) km depth that are —(5+9) km southeast of previously documented coordinates, between 18:34:52.0 ± 0.3 s, with latitude 40.724° ± 1.4 km and longitude 15.414° ± 1.4 km (preferred), and 18:34:52.5 ± 0.3 s, with latitude 40.742° ± 1.4 km and longitude 15.373° ± 1.4 km. The preferred nucleation point coincides with a downdip projection of the southeast end of the Carpineta fault and indicates that this fault ruptured first, rather than the Marzano fault as was previously thought.

With fault rupture nucleation point adjusted to this new preferred position, field and seismological estimates of seismic moment match well, both overall and for individual fault scarps, and suggest the following sequence of fault ruptures. The initial fault rupture nucleated at or near the southeast end of the Carpineta fault and propagated northwest, releasing ~2.5 X 10^18 Nm seismic moment. Rupture continued apparently without interruption onto the adjoining Marzano fault, where ~6.5 X1018 Nm of seismic moment was released. Rupture then paused for ~0.5 s, before continuing northwestward along the Picentini fault, where ~4.5 X 10^18 Nm more seismic moment was released. About 14 s after this sequence of NW-propagating ruptures began, a SE-propagating rupture released ~2X 10^18 Nm seismic moment on the San Gregorio fault. Each of these ruptures was associated with surface faulting and intense aftershock activity. The existence of another aftershock cluster northwest of the Picentini scarp suggests a fifth fault rupture, at Castelfranci, which released up to ~2 X10^18 Nm more seismic moment. Faulting at this locality ~12 s after the initial rupture began also appears necessary to explain the form of ground acceleration recorded nearby.

Two additional ruptures occurred on faults with different orientations, ~20 s and ~40 s after the initial rupture. The 40 s subevent involved the release of —3 X1018 Nm of seismic moment on a steep normal fault that dips southwest at ~70° and reaches the Earth's surface ~11 km northeast of the Marzano fault. The 20 s subevent apparently involved the release of ~4X 1018 Nm of seismic moment on a surface dipping northeast at ~20°, at the base of the brittle upper crust beneath this steep antithetic fault.

Points where ruptures nucleated on the steep NE-dipping normal faults coincide with en echelon steps of ~1 km and abrupt ~15° changes in strike. The Marzano and Carpineta faults, which have strike ~315°, took up a small component of left-lateral slip, as is revealed by the first-motion focal mechanism, teleseismic waveform modelling, striations measured in the field, and consistent rightward stepping: their slip vector azimuth is ~N37°E. Assuming the same slip vector azimuth, a component of right-lateral slip is expected on the San Gregorio and Picentini faults that have strike ~300°.