This paper describes a thick section of upper Miocene(?) to middle Pliocene marine sedimentary rocks exposed at Punta Perico, northeastern Carmen Island, in the southern Gulf of California. These strata overlie Miocene volcanic rocks along a lowangle unconformity and have a total thickness of ,1100 m. The section contains the following units: (1) lower conglomerate (,750 m thick) representing submarine debris¯ows and grain¯ows; (2) conglomerate and sandstone (,90±160 m) deposited by gravelly grain¯ows and sandy turbidity currents; (3) marlstone and mudstone (,120 m) with benthic forams that record bathyal water depths (400±500 m) and an age of ,3.5±3.1 Ma; (4) dacite breccia and boulder beds (,20 m) emplaced by submarine avalanches and debris¯ows and (5) strati®ed conglomerate and bioclastic limestone (^80 m) deposited by traction currents in a shallow shelf environment. Conjugate syn-sedimentary faults, deformation bands, and growth structures are recognized within a 250-m-wide belt adjacent to a large N-striking, W-dipping normal fault that bounds the eastern margin of the Pliocene section, indicating that the eastern normal fault was active during deposition of at least the upper ,300 m of the dipping section. Considering the thickness of the two units below the marlstone (,900 m) and assuming a range of possible sediment-accumulation rates (0.2±1.0 m/ka), we postulate that deposition of the lower conglomerate was initiated in late Miocene or early Pliocene time (,8±4 Ma), possibly coincident with initial opening of the Gulf of California. Deposition of the marlstone unit above a probable unconformity represents a major change in basinal tectonics at ,3.5±3.1 Ma, which is about the time when sea¯oor spreading and transform faulting were initiated in the Gulf of California. These tentative correlations suggest that the Baja California peninsula was structurally and mechanically linked to the deep Gulf during late Miocene to middle Pliocene time. q