To further our understanding of the paleolatitudes of Precambrian glaciations, a paleomagnetic study has been conducted on the glaciomarine Coleman Member of the Gowganda Formation and the conformably overlying deltaic Firstbrook Member and Lorrain Formation (Huronian Supergroup, ∼ 2.4-2.3 Ga) in Ontario, Canada. Many of the rocks, notably the gray and grayish red argillic facies, display unstable magnetization or overprints carried by magnetite and/or hematite that give negative fold tests. Some red sandstones, however, have stable high-temperature (∼ 690°C) components marked by very shallow northerly or southwesterly directions. Five sites in pale red sandstone from the Coleman Member near Elliot Lake give a dip-corrected A direction of D = 216.3°, I = 5.5° ( α 95 = 6.7 °, n = 30), and 17 sites in Lorrain red sandstone near Desbarats give a dip-corrected A direction of D = 5.4°, I = 5.5° ( α 95 = 5.9 °, n = 75). Unfavorable structural attitudes rendered fold tests inconclusive. These very shallow directions may be ascribed to early chemical remanent magnetization (CRM) because (a) some specimens display polarity reversals, (b) such directions are unknown, either as original or as overprint components, in post-Huronian rocks of the region, and (c) a prior site in Coleman red beds gave a high-temperature, shallow NNW direction that is not significantly different from Lorrain A and which a positive conglomerate test suggested is not an overprint. The very shallow directions are near primary directions for the slightly older (2.45 Ga) Matachewan dyke swarm in the region. Lorrain argillic rocks at five sites give a normal component B ( T = 670°C, D = 18.6°, I = 59.6°, α 95 = 6.1 °, n = 16, dip corrected) and a reverse component E ( T = 500°C, D = 190.9°, I = −60.9°, α 95 = 7.3 °, n = 14, in situ), which are close to a prefolding direction carried by magnetite previously identified in Coleman gray argillites (Morris A). However, a primary origin for Morris A is dubious because of the widespread occurrence of overprints carried by magnetite in Coleman gray argillites and because the fold test employed folds formed during the Penokean Orogeny at 1900-1800 Ma. Our results suggest that Morris A/Lorrain B and Lorrain E may data from immediately before and just after Penokean folding, respectively. Hence Coleman-sandstone A and Lorrain A may be better estimates of original directions. Those components and Matachewan paleomagnetic data constrain Huronian paleolatitudes and suggest that Huronian glaciation occurred within 11° and possibly within 4° of the paleoequator.