This comparison of the effects of LOGO use with the use of teacher-directed problem-solving instruction, and with conventional mathematics instruction, focused on the problem-solving ability, basic skills achievement, and attitudes of junior high school learners. Students (N=97) in five seventh grade mathematics classes were systematically assigned to three treatments: a problem-solving strategies instructional treatment that used printed worksheets, a structured LOGO treatment, and a control group. Learners were then assessed on their achievement, attitudes, and higher-level thinking skills using the Program Criterion Reference Test (PCRT), the Comprehensive Test of Basic Skills, the Revised Math Attitude Scale, the School Attitude Measure (SAM), the Test of Cognitive Skills (TCS), and the Test of Non-Routine Problem-Solving Skills. Results indicated that: (1) neither the LOGO group nor the problem-solving strategies group demonstrated any improvement in basic skills achievement as the result of the experimental intervention; (2) the problem-solving group scored significantly higher than the other two groups on both measures of problem-solving skills; and (3) while learners in the LOGO and problem-solving groups scored significantly higher than their counterparts in the control group in the Revised Math Attitude Scale, this can be, in part, attributable to a novelty effect. It is noted that the problem-solving skills fostered through LOGO use may not transfer outside the context of LOGO, since LOGO provides only a single algorithm which may not apply to many types of non-routine problems. A list of references, one graph, and 12 data tables are appended. (JB)