Neoproterozoic glacial records have been discovered on 23 palaeocontinents, their rate of discovery changing little since 1871. Yet, half of all the resulting publications appeared since 2000. The history of research before 1998 is described in five stages defined by publication spikes; subsequent work is not covered because historical perspective is lacking. In stage 1 (1871-1907), 'Cambrian' (now Neoproterozoic) glaciation was recognized successively in Scotland, Australia, India, Norway, Svalbard and China. Criteria for recognition included faceted and striated pebbles in matrix-supported conglomerates resting on ice-worn bedrock pavements. In stage 2 , Neoproterozoic glaciation was shown to have been widespread in Africa, Asia and the Americas. Major textbooks summarized these findings, but the rejection of continental drift (to account for late Palaeozoic glacial dynamics) put a chill on research. In stage 3 , the occurrence of glacial deposits within carbonate successions, as well as nascent palaeomagnetic observations, suggested that Neoproterozoic glaciers reached sea-level at low palaeolatitudes, but the belated recognition of sediment gravity flowage caused glacial interpretations to be prematurely abandoned in key areas. In stage 4 (1965In stage 4 ( -1981, the extent of Neoproterozoic glaciation was rethought in light of plate tectonics. Distinctive chemical sediments (iron + manganese formations and cap carbonates) were identified. In basic climate models, modest lowering of solar luminosity resulted in global glaciation due to ice-albedo feedback, and deglaciation due to greenhouse forcing resulted from silicate-weathering feedback in the carbon cycle. Neoproterozoic glacial geologists were blind to these ideas. In stage 5 (1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997), reliable palaeomagnetic data combined with glacial marine sedimentation models confirmed that Neoproterozoic ice sheets reached sea level close to the palaeoequator.