The effects of modeling and rehearsal in counseling nonassertive adolescents were examined by randomly assigning subjects (Ss) to one of five treatment groups: modeling plus rehearsal (MR), modeling only (M), rehearsal only (R), placebo control (P), or delayed-treatment control (C). Significant (p < .05) main effects for treatment were found on a self-report assertiveness questionnaire (AQ) and specific problem inventory (SPI). Newman-Keuls comparisons indicated that Post AQ scores of MR Ss were significantly (p < .05) more assertive than P or C Post AQ scores and that the MR, M, and R Post SPI ratings were significantly (P < .05) less severe than ratings in the P treatment group. (Author)