This study examines the effects of the computer-assisted text analysis program, WRITER'S WORKBENCH, on writing performance, levels of writing apprehension, students' writing processes and attitudes about using the computer and WORKBENCH for writing. A sample of 275 subjects enrolled in freshman composition were divided into an experimental group (N = 200) who used WORKBENCH in a mandatory computer lab component in addition to their composition course and a control group (N = 75) who received only the course, itself. Because random selection of participants was not possible, a Nonequivalent Control Group design was utilized. Holistic scoring of pre and posttest essays revealed a significant improvement in writing among both groups as a result of the treatments, but there was no significant difference in writing gains between the group using WORKBENCH and the group who did not (p = .942) . Similarly, though both groups demonstrated a small decrease in writing apprehension after instruction, there was no significant difference in the degree of decrease between the two groups (p = .201). Also, the data did not support a relationship between writing performance and apprehension. A 40 item questionnaire was given to the experimental group to determine: 1) attitudes about writing with a computer, 2) how students use WORKBENCH, and 3) students' attitudes about WORKBENCH. Some highlights of these findings are that narrow majorities enjoyed and were comfortable using the computer and WORKBENCH, but substantial minorities dissented or were uncertain. While 60% felt happier with their essays after using WORKBENCH and preferred using a computer to write, 89% of students felt word processing represented the greatest advantage and SPELL was the next most popular feature. Personal interviews conducted with 13 of the most and least apprehensive WORKBENCH users revealed that some students ignored the WORKBENCH analyses, and highly apprehensive students experienced more frustration with the computer, employed different writing processes, used WORKBENCH less often and less skillfully, and expressed more dissatisfaction with the computer.