Soils form on land surfaces by the actions of physical, chemical and biological processes on the lithosphere, and are influenced by climate, parent material, relief, organisms and duration of formation. Remnants of Plio-Pleistocene soils may be buried beneath younger deposits or persist on present land surfaces. Their potential for rigorously differentiating interglacials by climatic characteristics is limited by problems of: (i) precise dating of the beginning and end of soil-forming periods; (ii) distinguishing characteristics attributable to climatic factors from those related to parent material, relief, etc; (iii) calculating mathematical relations between measurable soil features and climatic variables; (iv) diagenetic changes in buried soils; (v) recognition and dating of relict features in unburied soils; (vi) loss of many soils by erosion. Some of these problems may be overcome if sequences of buried soils in periglacial loess deposits are used to compare the climates of successive interglacials in Europe and Asia. With the use of the length of interglacials derived from the oceanic record, the interglacials of the past million years are ranked according to approximate rate of soil development in loess. Two provisional equations relating soil development to time and climate are used; a linear relation probably overestimates the effect of time, and a logarithmic one seems to underestimate it. I tentatively suggest that oceanic oxygen-isotope stage 5e was warmer and wetter than the Holocene, stages 7 and 9 were cooler and drier than 5e, and 13-23 were generally warmer and wetter than 1-11.