Indigenous tussock grassland in New Zealand has a history of extensive pastoralism, and burning has been used to remove litter to improve establishment of aerially oversown pasture species and to promote palatable tussock growth for livestock. In recent years, considerable areas of tussock grassland have been retired from grazing and formally protected. Conservation land managers, as well as farmers, require information on the impacts of both managed burns carried out in spring and accidental fires that usually occur in warmer, drier conditions in summer. This study investigated the impact of spring and summer tussock grassland burning on the predominant soil microarthropods, Collembola and Acari, at 2 sites in Otago, in the South Island of New Zealand. Quantitative sampling was carried out before and for up to 26 months after burning replicated 1-ha plots. Total density of microarthropods in unburnt plots covered a similar range at both sites with an average over 3 years of about 18 000–20 000·m⁻² at each site. Both sites shared a dominance of Mesostigmata and Oribatida (Acari) and Isotomidae (Collembola). Burning in spring reduced densities of Oribatida after treatment at both sites for the duration of the study. However, after initial postburn reductions in density, populations of Isotomidae and Poduroidea (Collembola) recovered in the second year after burning. Prostigmata (Acari) appeared to be unaffected by fire. The effects of spring and summer grassland fires on microarthropod densities were rarely different. It was concluded that longer-term sampling would be required to observe the full recovery period for microarthropod populations after fire but that results from this study indicate rapid recovery of some microfaunal populations after fire, which is not strongly influenced by seasonal effects.