Experimental manipulations of soil fauna are a powerful tool for assessing causal relationships between belowground biodiversity and key ecosystem properties. However, preparing soil microcosm treatments without soil fauna for ecological experiments can be problematic. Methods to exclude nematodes, a ubiquitous and functionally important component of terrestrial ecosystems, have been developed for a few specific ecosystems, some of them involving the application of nematicides that may have interactive effects throughout the soil food web. Our goal was to develop a method to remove nematodes from soils of three Long Term Ecological Research (LTER) grassland sites, ranging from desert to moist, without use of chemicals and with moderate disturbance. Moreover, we aimed at testing whether the nematode removal would remain effective up to several weeks later. The following treatments were applied to ~3-kg soil microcosms in the laboratory: (1) a 72-hours heating (65°C) -freezing (-20°C) -heating (65°C) cycle using soil maintained at its original water content, and pre-wetting soil 24 hours before heating (65°C) for either (2A) 48 hours or (2B) 24 hours. We measured treatment effects on total abundance and trophic structure of the nematode community. To investigate whether nematodes would recolonize eight weeks after treatments, we conducted a greenhouse experiment where individual seedlings of the dominant grass species for each ecosystem were transplanted to treated and non-treated (control) soils. A heat-freeze-heat cycle of 72h using soil in its original field water content killed 60, 95, and 99% of the nematodes for the desert, semi-arid, and moist tallgrass prairie soils, respectively. Pre-wetting soil before heating increased mortality to 99% for all ecosystems after only 24h at 65°C.Root-feeders were the most resistant nematode trophic group. Eight weeks after treatments, there was no significant nematode recolonization for the pre-wetted 48h heated soils from the three sites, while for the semi-arid and moist sites there was a slight recovery in abundance in soil from the 24h heating treatment. Therefore, a treatment at 65°C for 48h using pre-wetted soil is recommended for eight-week long manipulative experiments in order to assure the effectiveness of the nematode removal throughout the experiment.