This study aimed to develop and evaluate multiple-representation-based student worksheets (LKPD) on colloidal system material and examine their effects on senior high school students’ cognitive learning outcomes and scientific attitudes. Using Thiagarajan’s 4D development model, which includes Define, Design, Develop, and Disseminate stages, the LKPD was designed, validated by expert lecturers, revised, and implemented with Class X students at SMA Negeri 2 Biak. A quasi-experimental posttest-only control group design was employed, with simple random sampling to assign experimental and control classes. Instruments included product practicality and feasibility assessments, cognitive learning tests, and scientific attitude questionnaires. Data were analyzed descriptively and inferentially using MANOVA and effect size calculations to determine the percentage of effective contribution. Results indicated that the LKPD was well-structured, printed in color on A4 paper, incorporated discovery learning syntax, and passed expert validation and readability tests. Students using the LKPD demonstrated significantly higher cognitive learning outcomes and scientific attitudes compared to the control group. MANOVA results revealed a very significant simultaneous effect of LKPD use on both variables (p < 0.001) with an overall effective contribution of 60.1%, a significant effect on cognitive learning outcomes (p < 0.001, 39.4%), and on scientific attitudes (p < 0.001, 54.3%). These findings suggest that multiple-representation-based LKPD is an effective instructional tool for enhancing both conceptual understanding and scientific attitudes in chemistry education.