The Late Paleozoic Ice Age (LPIA) lasted approximately 100 million years, starting in the Andean basins of western South America by the Famennian-earliest Tournaisian and culminating in eastern Australia by the Wuchiapingian. Based on a comprehensive dataset of the LPIA in Gondwana, we analyze how the trace-fossil record reveals temporal and spatial changes in high-latitude stressors reflecting the diachronous nature of the glaciation. Twelve trace-fossil associations, showing variable degrees of recurrence and named after one of their most characteristic ichnotaxon, were identified. These are the Umfolozia, Diplopodichnus, Mermia, Gyrochorte, Psammichnites, Planolites, Rosselia, Phycosiphon, Diplocraterion, Rhizocorallium, Zoophycos, and Thalassinoides associations. Almost all these associations illustrate archetypal Seilacherian ichnofacies. Salinity is regarded as a first-order environmental control, and associations can be further grouped accordingly. Other controlling factors that are envisaged as having played significant roles were hydrodynamic energy, substrate, sedimentation rate, and oxygen. The Umfolozia, Diplopodichnus, and Mermia associations are typical of freshwater settings, encompassing from proglacial lakes to fjord lakes and fjord heads. The Planolites, Gyrochorte, Psammichnites, Rosselia, and Phycosiphon associations reflect either permanent brackish-water conditions or fluctuations of marine or near-marine normal salinities and events of freshening, representing a wide variety of marginal-marine environments, such as estuarine, fjord, delta front, and prodelta. The Diplocraterion, Rhizocorallium, Zoophycos, and Thalassinoides associations are typical of normal-marine conditions, encompassing nearshore, offshore, and shelf environments, with the latter being characteristic of erosional exhumation of firmgrounds during transgressive ravinement. The fact that most of these associations illustrate archetypal Seilacherian ichnofacies shows that the basic tenets of the ichnofacies paradigm can be applied to these high-latitude Gondwana basins. However, there are some peculiarities of Gondwanan trace-fossil associations that need to be framed within the context of the LPIA. The volume of freshwater release due deglaciation in the Andean basins in the late Serpukhovian to Bashkirian and in the Parana and Karoo basins in the Gzhelian resulted in the displacement of the salinity barrier several tens of kilometers seaward within fjord systems. Freshening prevented the establishment of a marine benthos and allowed colonization by a freshwater biota, as indicated by the presence of the Umfolozia or Mermia associations. Marine ichnofaunas during during the LPIA show a very restricted distribution and very low levels of diversity, as shown by brackish-water suites identified in South America and India. The persistent low diversity of these marine trace-fossil suites suggests that stressful conditions continued even after glaciation in these regions. Overall climatic amelioration in Gondwana during the last stages and the aftermath of glaciation was signaled by a transition to optimal conditions for benthic life. Fully marine trace-fossil associations of these stages, as particularly illustrated in Australia, are essentially undistinguishable from those recorded in late Paleozoic basins of Laurasia. Proxies to assess trends in animal-substrate interactions may be seen as indicators of the environmental conditions that prevailed during an icehouse stage and its subsequent transition to a greenhouse time.