Eight palaeacanthocephalan species have been reported to modify crustacean intermediate host behavior in ways that may facilitate transmission of the parasite. These include members of three genera reviewed by Camp and Huizinga (1979, J. Parasitol. 65: 667-669), as well as Corynosoma constrictum in amphipods (Bethel and Holmes, 1973, J. Parasitol. 59: 945-956) and Plagiorhynchus cylindraceus in terrestrial isopods (Moore, Ecology 64: 1000). Recently, two species of eoacanthocephalans (Neoechinorhynchus cylindratus and Octospiniferoides chandleri) were shown to have similar effects on ostracods (DeMont and Corkum, 1982, J. Parasitol. 68: 125-130). Most of these animals induce increased activity and increased, positive photoreaction in their intermediate hosts. I therefore investigated the influence of an archiacanthocephalan, Moniliformis moniliformis, on the phototaxis and activity patterns of its cockroach intermediate host, Periplaneta americana. I also investigated response to air currents as a measure of predator avoidance (Silverman and Bell, 1979, Anim. Behav. 27: 652-657), shown to be important in P. americana escape behavior from toads (Camhi, Tom, and Volman, 1978, J. Comp. Physiol. 128: 203-212). Moniliformis moniliformis was obtained from wild Norway rats, Rattus norvegicus, collected along the Houston (Texas) Ship Channel in January 1982. Adult cockroaches were held for 1 wk without food and then given a mixture of applesauce and shelled acanthors taken from gravid female worms. Control cockroaches were treated similarly, but were not given acanthors. Cockroaches were maintained on rat chow and water at 25 C, 12L: 12D, scotophase beginning at 2000 hr. At 23 to 26 C, cystacanths of M. moniliformis in cockroaches become infective in 7 to 8 wk PI (Moore, 1946, J. Parasitol. 32: 257271). Thus I allowed a minimum of 2 mo to elapse before using infected cockroaches in experiments. Because Roberts (1960, J. Cell. Comp. Physiol. 55: 99-110) reported that activity patterns in female P. americana were more erratic than those of males, I studied only male cockroaches. Preliminary work with infected females was inconclusive (Moore and Romo, pers. observ.). At the termination of all experiments, cockroaches were dissected and examined for cystacanths. Where scores are reported, SE follows the mean. The black behavioral arena (diameter = 1.22 m) in which phototaxis and air current experiments were conducted was surrounded by walls 35.6 cm high and rimmed with Fluon and petroleum jelly to prevent escape. Three, red, 25watt bulbs furnished ambient light. Prior to these experiments, which were conducted for no more than 2 hr at the beginning of the scotophase, samples of infected and control cockroaches were mixed. They were then tested randomly. At the beginning of photoresponse experiments, I placed a cockroach under a black pan in the middle of the arena. A 75-watt bulb illuminated the arena through a 232-cm2 screened opening in the wall. Alternately, two such openings were used so that response would not be a function of direction. After 5 min, the pan was lifted, exposing the cockroach to directional light. I recorded the absence or duration (sec) of a stationary (freezing) response and the direction of movement (0 = directly toward the light, 6 [maximum score] = directly away from light, etc.). In the air current tests, the animal was allowed to acclimate for 5 min, and then air was quickly discharged from a 10-cc syringe approximately 6 cm from and perpendicular to the cockroach dorsum (Silverman and Bell, 1979, loc. cit.). Ensuing behavior was scored for distance moved (cm), absence or duration (sec) of the freezing response, and time spent on the vertical surface of the are-