BACKGROUND. Patients with dense ocular media disorders retain the ability to project or point to an intense source of light. Using this response capability and high luminance points of light as stimuli, Vernier judgments (a hyperacuity test) can be made by these patients, even without the presence of a "window" through a leucoma, cataract, or bleed. Without coaching, these individuals are able to locate the centers of the individual degraded point images if the individual light sources are adequately separated (i.e., if sufficient "gaps" exist between the individual stimuli), and they can spatially align the degraded images. Advanced cataracts are the main cause of blindness in the developing world, and this is a treatable condition. In these nations, only a modest proportion of affected patients receive surgery, and only 5% or less of these individuals obtain treatment in two eyes. There are incredibly large and rapidly growing backlogs of advanced cataract patients requiring care (many millions). Because of the 20 to 30% failure rates that occur after treatment (all causes) in many developing world settings, a test performed before surgery, which offers a meaningful estimate of postsurgical visual outcome, can be valuable. Using the principle defined above, we seek to determine before surgery those individuals who will derive most benefit from cataract removal, and which of two cataractous eyes has the better postsurgical visual prognosis. EXPERIMENTAL. In Berkeley, we performed a series of preliminary studies on a Vernier acuity test before initiating a clinical study in a developing world setting. These studies were conducted upon young adult normal subjects wearing their usual vision corrections, with and without induced refractive errors, and/or with or without simulated dense nuclear cataracts. We sought (1) to determine the number of repeat trials necessary for reliable outcomes; (2) to compare a two-point and a three-point Vernier acuity display; (3) to determine the shape of the measured response function at large gap separations between test points; (4) to define optimal test distance and stimulus size; (5) to assess the effect(s) of a broad range of uncorrected refractive errors upon outcomes; and (6) to consider means to minimize refraction-based errors by using a pinhole, a refractive correction, and/or selective spatial filtering. We compared responses obtained using the current CRT/VDT-based, computer-driven (Berkeley) instrument with a new precision optical/mechanical computer-driven (India) instrument. The India instrument is needed to determine design parameters for a next stage simpler, cheaper, more rugged field instrument(s).(ABSTRACT TRUNCATED AT 400 WORDS)