Visually induced motion sickness, or “cybersickness”, has been well documented in all kinds of vehicular simulators and in many virtual environments. It probably occurs in all virtual environments. Cybersickness has many known determinants, including (a short list) field-of-view, flicker, transport delays, duration of exposure, gender, and susceptibility to motion sickness. Since many of these determinants can be controlled, a major objective in designing virtual environments is to hold cybersickness below a specified level a specified proportion of the time. More than 20 years ago C. W. Simon presented a research strategy based on fractional factorial experiments that was capable in principle of realizing this objective. With one notable exception, however, this strategy was not adopted by the human factors community. The main reason was that implementing Simon's strategy was a major undertaking, very time-consuming, and very costly. In addition, many investigators were not satisfied that Simon had adequately addressed issues of statistical reliability. The present paper proposes a modified Simonian approach to the same objective (holding cybersickness below specified standards) with some loss in the range of application but a greatly reduced commitment of resources.