Entomopathogenic nematodes are important biological control agents for a variety of soil- and litter-dwelling insect pests. A major drawback to their use against pest species is their low level of persistence in many agricultural systems. While a number of studies have examined the persistence of these biological control agents over periods of days and/or weeks, the longer-term survival of these nematodes has received less attention. We report the results of a year-long field experiment testing the long-term survival of infective juveniles (IJs) of the entomopathogenic nematode Heterorhabditis marelatus (Liu and Berry), under a range of initial densities. We buried mesh-covered tubes containing raw field soil with varying densities of H. marelatus IJs in a coastal prairie containing naturally occurring populations of this nematode and then destructively sampled subsets of the tubes for nematode presence five times over 1 yr. H. marelatus IJs lived a surprisingly long time in the absence of prey: some nematodes from the initial cohort were viable after a year in the field. Survival over the year-long course of the experiment was independent of the starting IJ density, suggesting that H. marelatus does not aggregate at high densities to reduce desiccation risk. Our results highlight the fact that IJs vary greatly in their long-term survival; selecting entomopathogenic nematode isolates for persistence as well as virulence could enhance this biological control agent’s long-term effectiveness in agricultural systems.