The Permo-Pennsylvanian Cutler Formation in the proximal Paradox basin (western equatorial Pangea) has long been interpreted as low-latitude alluvial-fan deposits shed from the Uncompahgre uplift during the Ancestral Rocky Mountains orogeny. Here, we employ sedimentologic analysis of the Cutler strata within 10 km of its exposed source to hypothesize the alternative interpretation of deposition in a tropical proglacial-periglacial system. At this proximal location, the Cutler Formation onlaps Precambrian basement and dips gently southwestward, such that the section youngs distally. Along the onlap contact, crudely bedded and subhorizontal granule to cobble conglomerate of traction-flow origin is angularly juxtaposed atop inclined (15-30°) boulder diamictite and granule conglomerate of mass-flow origin (inferred subaqueous mass flows). The steeper dips form a semi-radial pattern that, together with facies evidence, suggest a subaqueous fan near this location. Immediately basinward (1-3 km from the onlap) is a gently dipping (≤ 10°) assemblage of cobble diamictite, granule conglomerate, sandstone, and mudstone of inferred subaqueous mass-flow origin. We propose that these associations collectively record topset, foreset and proximal bottomset deposits, respectively, of a lacustrine Gilbert-type delta. The slightly younger Cutler Formation exposed from 3 to 10 km from the onlap contains subhorizontal (≤ 4-7°) granule conglomerate and subordinate sandstone deposited by hyperconcentrated flood flows and traction flow in a fluvial system characterized by frequent high-discharge flow. Scattered siltstone units form a distinct assemblage reflecting silt genesis and eolian sorting (loess deposits), but with common overprinting by lacustrine and fluvial processes. We propose that these facies document the presence of generally abundant water. However, chemical and mineralogical immaturity record relatively minimal chemical weathering most compatible with a cold and wet system. Lateral to the inferred topset exposure, the Cutler strata fill a paleotrough within the Precambrian basement and contain abundant lonestones interpreted to record ice rafting, suggesting that the lacustrine system was, at times, ice-contact. Abundant faceted clasts and loess are also consistent with a glacial influence. Ultimately, the onlap of the Cutler strata onto the Precambrian basement of ancestral Unaweep Canyon, which has elsewhere been hypothesized to be a late Paleozoic valley of glacial origin, establishes a potential link between the two. We propose that, as an alternative to the long-held view of an alluvial fan in a hot (tropical) climate, the facies and climate indicators of the proximal Cutler system record a proglacial to periglacial, largely lacustrine-fluvial system related to tropical glaciation of the Uncompahgre uplift during the late Paleozoic. The proximity of marine and paralic facies to the Cutler "fan" system further implies that such glaciation resulted from cold temperatures, rather than high elevations, thus indicating a globally cool planet during the time recorded by the study section. If valid, this hypothesis raises the possibility that upper Paleozoic strata in other, uplift-proximal equatorial systems record anomalously cold conditions, and perhaps merit reassessment.