This publication presents the results of research involving different types of self-cementing fly ashes (without any other activators) for the stabilization of four different types of soft subgrades from various road sites in Wisconsin, USA. The strength approaches were applied to estimate the optimum mixture design and to determine the thickness of the stabilized layer. The stabilized soil samples were prepared by mixing fly ash at different contents at varying water contents. The performance of fly ash stabilized subbase depends both on the specific source of fly ash and the engineering properties of soils. It is suggested that the performance analysis of fly ash should be based upon the laboratory tests such as index properties, compaction, unconfined compressive strength and CBR tests of a specific site. This is suggested rather than using the study of the physical properties and chemical composition of fly ash and soil. As disclosed in the literature, the strength gain due to stabilization depends mainly upon three factors; ash content, molding water content and compaction delay. The samples were subjected to unconfined compression strength and California bearing ratio (CBR) tests after 7 days curing time to develop water content-strength relationship. To evaluate the impact of compaction delay that commonly occurs in the field during construction, the sets of samples were compacted 2 h later after mixing with water. The unconfined compression strength and CBR tests were performed and used to determine the thickness of the stabilized layer in pavement design. All of these factors were taken into account throughout this research. © 2005 Elsevier B.V. All rights reserved.