This thesis contains results from three projects describing self-cementing fly ash stabilization of RAP-soil mixtures, stabilization of limestone screenings for use as a structural layer in road construction, and finite element modeling results of various subgrade materials including self-cementing fly ash stabilized subgrade, natural subgrade, granular subbase, and hydrated fly ash. The first project shows that self-cementing fly ash stabilization of RAP-soil mixtures is economically feasible and structurally capable of supporting construction traffic. The increase stiffness from the addition of self-cementing fly ash increases capacity ensuring long term pavement performance. Addition of self-cementing fly ash increases the consolidated shear strength about five times. The second project shows construction operations and field results proving that stabilization of limestone screenings is viable, cost effective, and produces an adequate structural layer for road construction. The measured moisture-density curves for manufactured sand and limestone screenings are about the same, and the moisture-strength curves show a dramatic decrease in strength beyond the optimum moisture content for strength. Durability testing concluded that CKD stabilized manufactured sand and limestone screenings are not viable construction alternatives, and the addition of class C fly ash with CKD significantly increased the durability of the mixtures. The third project concluded that a link exists between subgrade non-uniformity and pavement performance. Field testing, with the DCP, Clegg Impact Hammer, nuclear density gauge, and GeoGauge, and statistical analysis of subgrade materials concluded that granular subbase, self-cementing fly ash treated subgrade, and HFA decrease the variability of field results. Finite element modeling analysis proved that a link exists between subgrade non-uniformity and pavement performance. Uniform modeling conditions produced lower average deflections and stresses increasing pavement life. Statistical analysis concluded that modeling uniform subgrade conditions produce average stresses that have less variability than those for non-uniform modeling conditions. Pavement response reliability increased with the addition of uniform subgrade, proving that subgrade non-uniformity influences pavement performance.