A program of construction and the performance evaluation of three major Virginia experimental flexible pavements is reported. The objective of the program was to evaluate the performance of pavements incorporating new or timely design concepts and to assess the feasibility of these concepts for further use. Among the major findings of the study are: 1. A resilient select borrow material used over a resilient subgrade does not enhance pavement performance. When used between a cement stabilized subgrade and a crushed stone base, the resilient select material may, in fact, impair performance. There is some evidence that resiliency is reduced after a substantial number of wheel loads, with an improvement in subsequent pavement performance. 2. Pavements having cement treated crushed stone under a thin (3 in.) bituminous structure have performed very poorly. 3. Transverse shrinkage cracks reflect from a cement treated stone subbase through 3 inches of bituminous concrete in as little as 18 months, and through 7 inches in less than 5 years. 4. Where cement stabilized stone subbases are used and when truck traffic is normally channeled into the outer lanes, advantage may be taken of the omission of the stabilization from stone subbases under inner or passing lanes. 5. Shrinkage cracking aside, 4 inches of cement treated aggregate base, 4 inches of bituminous concrete, and 6 inches of untreated crushed stone base (all underlying 7 inches of bituminous concrete) give approximately equal performance after 5 years under heavy traffic conditions. Higher deflections and greater cracking on the pavement incorporating the untreated stone suggest that the performance may not be equal after many load applications.