The Paleoproterozoic Piling Group along the southeastern Rae margin, northern Canada, is characterized by thick, deep marine turbidite deposits not observed in time-equivalent, intracratonic basin units further southwest. Models invoked to explain this feature include development of a full-ocean, back-arc, or proto-ocean basin followed by turbidite sedimentation. We present new and existing U-Pb geochronological, Nd isotope, geochemical, and stratigraphic evidence that support a proto-ocean basin model, and we explore the events leading to the formation and closure of such a rift basin during the middle Paleoproterozoic. Sedimentation initiated largely after ca. 2160 Ma with deposition of craton-derived, shallow marine siliciclastic strata (Dewar Lakes formation). Continued extension resulted in accumulation of south-facing carbonate beds (Flint Lake formation) and likely concomitant, arc-like tholeiitic to picritic volcanism and voluminous volcaniclastic sedimentation (lower Bravo Lake formation) farther outboard at ca. 1980 Ma. Accumulation of intrabasinal siliciclastic strata above lower Bravo Lake formation rocks may mark a hiatus in mafic-ultramafic magmatism. By ca. 1923 Ma, upper Bravo Lake formation, within-plate–type alkaline sill emplacement and volcanism occurred within highly extended crust. The oceanic island basalt–like signatures of the Bravo Lake formation rocks (but lack of depleted, mid-ocean-ridge basalt–type compositions) suggest that by this time the thinned Rae continental lithosphere had fragmented into small crustal block(s) and narrow zone(s) of incipient oceanic crust farther outboard of the Piling Group basin. Rapid subsidence of the southeastern Rae margin ensued, leading to deposition of euxinic (Astarte River formation) and overlying turbiditic strata (Longstaff Bluff formation). The post–ca. 1915 Ma northern turbiditic sedimentary units were likely derived from a thoroughly mixed, two-component source with possible input from the Snowbird tectonic zone and Bravo Lake formation, whereas the post–ca. 1930 Ma southern turbidite unit may have been sourced from the Meta Incognita microcontinent, presently exposed further south. We favor a rift margin over a foreland basin setting for the deposition of the northern turbidite deposits. Subsequent mantle upwelling associated with incipient ocean formation may have triggered melting of highly thinned continental crust resulting in emplacement of late-stage, ca. 1897 Ma, contaminated rapakivi granite and highly differentiated mafic sills. Our results are most consistent, albeit not exclusively, with the much debated model of asthenospheric upwelling and incipient rifting along the Rae-Hearne boundary farther southwest at ca. 1.9 Ga. Later accretion of the Meta Incognita microcontinent led to basin closure and development of a north-verging fold-and-thrust belt after ca. 1883 Ma.