Two processing abilities used to solve spatial problems are examined: (1) the analog ability of structural visualization; and (2) the non-analog ability of verbal analytic reasoning. The distinction is based on an evaluation of information processing theory and a review of process-oriented studies of individual differences. Criteria are presented for determining which abilities are measured by tests of spatial ability and for classifying existing instruments on the basis of these criteria. Non-analog ability is the ability measured by certain tests of general intelligence and verbal processing abilities. Analog ability involves the holistic gestalt-like processing of visuospatial information. Reexamining the literature on individual differences and classifying the results by test types demonstrates that the commonly accepted belief that sex differences in spatial ability emerge in adolescence is not supported by the literature. The male advantage in analog processing ability is evident well before adolescence. Accumulated evidence also does not support the theory of recessive X-linkage of spatial ability. Five figures illustrate score differences, and 30 examples of spatial test items are appended. A 157-item list of references is included. (SLD)