Two training procedures (active rehearsal, passive observation) were assessed for relative effectiveness in the acquisition of sequential fire emergency skills, reducing fire-related fears and physiological arousal, and increasing self-efficacy and rationale for fire safety skills. Active subjects imitated videotaped models performing emergency behavior while receiving behavior-contingent feedback. Passive viewers received no practice or feedback. Experimental groups were compared to untrained controls. Active rehearsal was expected to lead to superior skill acquisition, fear-reduction, self-efficacy appraisals, rationale acquisition, and reduction of physiological reactivity (systolic and diastolic blood pressure, pulse rate). Subjects were 52 third-grade children. Dependent measures were assessed at pre-test, post-test, and five-month follow-up. Significant performance gains were found for both experimental groups, but not for controls. Active training produced significantly greater skill acquisition. Skill gains were not maintained at follow-up. All groups showed significant reductions in fear at post-test, but no differences were found between groups. Active training also produced significantly greater self-efficacy appraisals for one emergency situation. Both experimental groups demonstrated significant gains in rationale acquisition relative to controls. Physiological results were confounded by pre-test differences. Correlations between various measures were examined. A significant relationship was found between self-efficacy appraisals and behavioral performance for one emergency situation. The correlation between self-report of fear and self-efficacy was highly significant across time. The implications of active training strategies in fire safety programs are discussed. / Master of Science / incomplete_metadata