New mapping and Ar/Ar ages from volcanic rocks near Bahía de Guadalupe in the N-central Gulf extensional province (GEP) show that extension near the SE end of the Ballenas transform system began in late Oligocene or early Miocene time, and was episodic(?) into Quaternary time. Strata S and SW of Bahía de Guadalupe comprise 4 sequences: (1) basal, basement- and arc-derived sandstone and conglomerate of fluvial, eolian, and debris-flow origin (~24 Ma), (2) a silicic welded tuff, (3) basalt and basaltic andesite, and (4) Plio-Quaternary fluvial and shallow marine sediments. Sequence 1 is typically tilted E or NE 20-30° and commonly is more volcanogenic upward, with thin reworked tuffs (one dated at 23.97±0.33 Ma, hornblende), andesite-clasts, growth sequences (fanning dips, locally to ~70°), and internal angular unconformities. Soft, hornblende-bearing pumice clasts from gently dipping beds above one such sequence or unconformity yield an age of 24.51±0.28 Ma. The welded tuff has biotite (14.37 ±0.13 Ma) and plagioclase, is only locally preserved in paleovalleys, and may be correlative with deposits to the NW. Sequence 3 lies on basement, is locally subparallel to the welded tuff, or is in angular unconformity on sequence 1; it overlaps normal faults and fault-related paleotopography, but is also faulted. A sequence of olivine-pyroxene basalt yielded ages of 12.8±0.37 and 13.29±0.15 Ma. These flows are overlain locally by hornblende basaltic andesite, then by olivine basalt that gave an age of 11.8±0.5 Ma. Subsequent erosion left sequences 1–3 preserved above valleys that were partly filled with sequence 4 beds, which are flat lying in western exposures but tilted, faulted, and intruded near the coast and Ballenas transform.

These relationships show that initial extension in this part of the peninsula was unusually old (late Oligocene?-early Miocene), was coeval with arc volcanism, and suggest that rift segmentation into tilt domains is locally as old also. Old extension also occurred to the S (S of Bahía de Los Angeles) and farther N (Arroyo Candeleros). The area restores S or SW of Guaymas when the Gulf is closed, indicating that “Sonoran-style” extension reached anomalously close to the coast at that latitude, and may have affected the location of the SE ends of the Ballenas and Tiburón transform fault systems.