Recent developments in low-cost, large scale production of entomopathogenic nematodes of the genera Steinernema and Heterorhabditis (Bedding 1984) have facilitated large-scale field testing against various soil-infesting insects (Kaya 1985, Poinar 1986), with good potential for control in some applications. Early research by Poole (1976) and Quattlebaum (1980) indicated that the potential included fire ants. Citrus growers in Florida using nematodes (BioVectorg; Biosys, Inc.) for control of soil insects have reported fewer problems with fire ants (personal communications to W. R. Martin). Jouvenaz et al. (1990), however, did not achieve significant control of red imported fire ants, Solenopsis invicta Buren, in field trials of nematodes (Biosys, Inc.) conducted in Florida. Steinernematid nematodes may be moderately effective against fire ants in irrigated soil. Morris et al. (1990) reported 47% control with Steinernema carpocapsae versus 39% control with the Amdro standard at sites containing irrigated turf. This level of control is far below normal for Amdro (Williams et al. 1987; Jouvenaz et al. 1990), indicating that it may have been broadcast on wet soil (personal communication, D. F. Williams, USDA, ARS). Unmarked mounds found outside the blocks after treatment were not included in the assessment for activity; thus, control by nematodes may have been overestimated due to mound relocation. Considerable relocation of fire ant nests occurred in the tests conducted by Jouvenaz et al. (1990). Had not mound movement been taken into consideration (Poole and Quattlebaum scored all mounds uninhabited after treatment as dead), the apparent control would have been over 77% (of 44 treated mounds, 34 were inactive). We subsequently confirmed the aversion of fire ants to nematodes by controlling the movement of small laboratory colonies between containers of sandy soil with single applications of BioSafe? Lawn and Garden Insect Control (Steinernema carpocapsae; Biosys, Inc., Palo Alto, CA). The readiness with which fire ants relocated their nests in the field and vacated soil in containers to avoid nematodes prompted us to evaluate the potential of these parasites to eliminate fire ants from nursery stock. The early spread of fire ants in the southeastern United States was greatly facilitated by the shipment of infested nursery stock (Lofgren 1986). Recently these pests have been transported to Arizona and California; if they become established in the West, their range will increase substantially. Currently, federal certification of nursery stock for shipment through quarantine is based solely on incorporation of chlorpyrifos. Granular chlorpyrifos is not as effective as previously believed, drenching poses problems of worker exposure, and both treatments are expensive (personal communication, H. Collins, USDA, APHIS). The improvement of methods for control of fire ants is a critical need of the nursery industry (Regelbrugge 1991). In the present study, three sequential tests were conducted using BioSafe against fire ants nesting in nursery pots containing Pittosporum sp. shrubs. A colony of S. invicta consisting of five queens, 3,000-5,000 workers and about 1.5 g brood was introduced into each of 50 1 gal pots containing shrubs 20-35 cm in height. The ants were