Oxisols cover approximately 23% of the land surface in the tropics and support zero- to high-input agricultural systems in many tropical countries. Together with the closely related Ultisols, recent estimates indicate that these soils occupy from 38% to 46% of the tropics. Some of these soils still appear to have classification problems. This subsequently, appears to perturb successful implementation of sustainable agriculture. Since the introduction of Soil Taxonomy in 1975, several changes have been made to the definition of Oxisols. The kandi concept, for instance, was introduced to improve the classification of transitional soils. However, the definition for Oxisols in Soil Taxonomy (1999) indicates that additional input is still required to refine the definition in order to resolve some of the outstanding classification problems. In his forward message, H. Smith, the director of soil survey division has mentioned in a positive note, that despite many improvements in the second edition of Soil Taxonomy, there are still areas that will require a concerted effort to improve soil classification. Therefore, the objective of this study is to examine the properties and behavior of some Oxisols and closely related soils in order to evaluate their classification. Soils from Brazil, several countries in Africa and Malaysia were used in this study. Field observations on these soils provided the first indication that some of the presently classified kandi-Alfisols and kandi-Ultisols were closer to Oxisols in terms of their properties. Water retention differences and apparent CEC of the subsurface horizons also supported this idea. The types of extractable Fe-oxides and external specific surface areas of the clay fractions showed that many kandic horizons have surface properties that are similar to the oxic horizons. Micromorphology indicated that the genetic transition from the argillic to the oxic involves a diminishing expression of the argillic. The properties of kandic horizon is therefore, inferred as transitional between the argillic and the oxic horizon. In Malaysia, the Oxisols and some kandi-Ultisols have comparable FFB yields and response to fertilizer treatments. These soils are grouped under one soil management category. As Soil Taxonomy is a basic system of soil classification for making and interpreting soil surveys, the problem with the current definition for Oxisols is evident. Agro-management and yield data appear also to support the need to revise the definition for Oxisols. Oxisols, by concept, are soils that represent extreme manifestation of the soil-forming processes. Therefore, the processes that create this extreme manifestation should be used to define such soils at the soil order level. It is proposed that the Oxisols be keyed out based only on the presence of an oxic horizon. The definition for the oxic horizon has been revised and the clay increase requirement of item 5 of the former definition has been deleted. The impact of this proposal on Oxisols is that, soils with kandic horizons that meet the properties of the proposed oxic subsurface horizon, will still be grouped under this soil order. However, the presence of a kandic subsurface horizon would be reflected at a great group level. A textural differentiation with depth does not necessarily imply clay illuviation. Therefore, it would appear incorrect to emphasize the clay increase and as such, the presence of a kandic horizon, at the order level. There will still be some kandi-Ultisols or kandi-Alfisols; however, these are soils that have much higher apparent CEC. Testing of the proposed classification on some Malaysian soils for the last 10 years showed that the new definition for Oxisols provides a better basis for the classification of local soils and the development of meaningful soil management group maps for plantations to implement sustainable agro-management practices in the 21st century. This redefinition for Oxisols brings the classification of highly weathered soils closer to the 4th, 7th and 8th attributes of Soil Taxonomy (1999).