The Early–Mid Ordovician has long been considered a super-greenhouse world, based largely on high relative global sea levels and light stable oxygen isotope data from bulk carbonates. An alternative and largely untested hypothesis has suggested that, at least in equatorial palaeolatitudes, there was a steady cooling trend through the Early Ordovician reaching the range of modern equatorial sea surface temperatures by the Mid Ordovician. This hypothesis, though controversial, is supported by palaeobiogeographical studies of the early Late Ordovician (Sandbian) zooplankton (graptolite) biotopes. These are comparable to the modern planktonic foraminifera ‘provinces’ and suggest a latitudinal temperature gradient that is similar to the modern one.Chitinozoans are also an important group of Palaeozoic marine microfossils, but with a poorly known biological affinity. Here we present an analysis of chitinozoan species palaeobiogeography for the early Late Ordovician (Sandbian c. 460 Ma), which confirms that these microfossils, and likely their parent organisms, were epipelagic. Unlike the graptolites, chitinozoans had their highest diversity and abundance south of ~ 35°S during this time, which strongly suggests adaptation to temperate to cold waters. The distribution of chitinozoan biotopes during the gracilis time-slice allows us to identify the austral Subtropical, Subpolar and Polar belts; key oceanographic boundaries are identified as the Subtropical–Subpolar Transition at ~ 35°S and tentatively, the Polar Front at ~ 55°–70°S. Again, these are in a similar position to those in the modern Southern Ocean. Chitinozoan biotopes also map a steep latitudinal faunal gradient that is comparable to that of the graptolites and that of modern plankton. This likely indicates a steep latitudinal temperature gradient for early Late Ordovician seas and questions the notion that this interval of Earth history had a greenhouse climate.