The loess-paleosol sequences in New Zealand are among the thickest in the southern hemisphere, and preserve an important terrestrial record of paleoclimatic changes. Unlike loess-paleosol sequences in North Island, most of those in South Island lack discrete chronostratigraphic marker ash beds, excepting perhaps the widespread ~24ka Kawakawa Tephra. Therefore, any specific correlation to marine isotope stage (MIS) proxy records or to loess-paleosol sequences in the northern hemisphere has been unknown. Hence, we applied luminescence sediment dating to four key loess-paleosol sequences from southern South Island.Although these pioneering results suggest that further sampling and luminescence dating are needed, they do provide: (1) the first numeric age constraint (44+/-3ka on overlying loess) for the end of deposition of the geomorphically and paleoclimatically important Edendale Terrace gravels; (2) the first numeric evidence that the youngest loess-paleosol unit (L1) at the near-coastal Romahapa and inland Stewarts Claim sites began to be preserved much earlier (at ~60ka) than at the inland Kingston Crossing site (~35ka); (3) the first direct evidence that the 6m Romahapa sequence extends at least to ~350ka, the oldest yet TL-dated such sequence in South Island. Furthermore, the stratigraphically self-consistent (with one exception) luminescence ages between ~60 and ~350ka conflict with a previous indirect chronostratigraphy inferred from trace concentrations of glass shards. This trace-shard chronology implies a major hiatus in loess-paleosol preservation (missing at least MIS 6-7) not recognized by luminescence dating.