Six test drivers were tested over four nights each for their response to experimental treatment of rural intersections. The treatments were applied to four geometrically similar sites, two of which were State Route-State Route intersections and two which were State Route-County Route intersections. Treatments included use of one mercury, two mercury and two sodium luminaires and the use of special intersection delineators and signing (use of a W-24 sign on the approaches to the SR-CR intersections). Pavement marking (new versus old) was a third major treatment factor. Subjects approached the intersection along the state route with instructions to execute a left turn at the intersection. There were 168 approaches made at the test sites. Performance measures included the distance from the intersection at which a) the gas pedal was released, b) 35 mph was reached, and c) longitudinal acceleration was maximum, and d) 12 measures derived from eye-movement recordings. The major results were that when compared to a baseline, no-treatment condition, the use of lighting significantly improved performance with earlier detection of the intersection by the driver and subsequent smoother velocity profiles. Signing and delineators had only marginal effects and new pavement markings showed no effect. Current signing for SR-SR intersections was found to enhance performance over the SR-CR sites under no-treatment conditions.