Attempts to relate information-processing capacities to intelligence test scores have had modest success when normal subjects are used but have been very successful in showing differences between extreme groups in information-processing capacity. This is illustrated by reviewing studies relating memory to verbal comprehension. The reason for these disparate results may be that in normal subjects individual competence may depend largely on differences in choice of a problem-solving strategy. Using sentence verification as an example, it is shown that information-processing and psychometric measures are in much closer correspondence when account is taken of one's problem-solving strategy. This, however, still leaves us with the problem of explaining positive manifold, i.e. the psychometric evidence for a concept of general intelligence. The relation between g and the idea of attentional resources is explored, by applying the dual task methodology to memory and reasoning tests. The results support the notion that g can be related to attentional resources.