Two assumptions and seven hypotheses defining Fiedler and Garcia's (1987) cognitive resource theory are critiqued and examined with data from 48 four-man groups. Information for leaders and subordinates was available on the dimensions of IQ, directiveness, group atmosphere, rank, and group performance on seven tasks. Results yielded support for Hypothesis 2 of the theory (i.e., leader intelligence was more strongly correlated with group performance for highly directive leaders, relative to nondirective leaders). Problems with the theory associated with (a) the definition of intelligence, (b) the undervalued importance of specific task demands and the combinations of tasks, and (c) the role of stress are discussed.