A large part of our knowledge about the perception of heading has been accumulated using a passive observer paradigm. Recent experimental tasks have mostly been designed such that observers were presented with a simulation of being transported through an environment, akin to being passengers in a moving vehicle. Depending on type of simulation technique and simulated array, reported mean heading thresholds range from approximately 1–2° to 10°. Investigated velocities mostly fall into the range of human bipedal locomotion, that is, between 0.6 and 2.4 eye-heights per second (h/s). Warren, Morris, and Kalish found some support for the notion that increased translational speed results in greater heading accuracy. The data obtained by Warren et al. were more equivocal. The study was designed to evaluate the utility of an active control paradigm for the study of heading perception for two different simulation techniques, and the effect of speed of simulated self-motion on heading perception.