In the present research, it was tried to unravel the influence of various types of instruction on test anxiety levels and, in turn, its influence on intelligence test performance. Three types of instruction were compared: a stressful, achievement-orientated instruction; a reassuring, task-orientated instruction; and an ambiguous instruction. Before the instruction, trait anxiety was measured with the trait version of the state-trait anxiety questionnaire and negative fear of failure with a standard questionnaire. Within each instructional condition, there were two test sessions with four intelligence tests each. Before each session, state anxiety was measured with the state version of the state-trait anxiety questionnaire and test anxiety with a standard test anxiety questionnaire. It turns out that state anxiety and test anxiety do not increase more rapidly as a function of anxiety disposition under stressful conditions compared with reassuring conditions. As in previous research, stage-fright effects were observed in the sense that repeated measures of state anxiety and test anxiety showed a decline of average anxiety levels.