<![CDATA[The lack of availability of assessment instruments to measure cognitive ability at the C4-C6 ability level, the lack of giving practice questions at the C4-C6 level in Mathematics learning resulted in the assessment being carried out less than optimal. The research objective is to test the quality of the HOTS-oriented mathematics learning outcomes assessment instrument. This research is research on the development of 4D models by Thiagarajan. The stages are: define, design, develop, and disseminate. The research carried out only reached the development stage due to limited time, cost, and manpower. The research subjects were lattices and learning outcomes tests in multiple-choice as many as 20 questions. Data collection methods are test and non-test methods. The data collection instruments were judges' questionnaire sheets and learning outcomes tests. Content validity data analysis method used the Gregory formula, and item validity used the Product moment formula, discriminatory power, difficulty level, distractor quality, and reliability uses the KR-20 formula. The results of data analysis obtained a content validity test of 1.00, very high criteria. It has a distinguishing power of 16 items with the criteria of "good enough," 2 with the criteria of "good." It has a difficulty index of 1 item with "difficult" criteria, 14 items with "medium" criteria, and three items with "easy" criteria. Has a distracting quality of P > 5% as many as 53, P < 5% as many as 7. With a reliability test of 0.93, the category is very high. Based on the results of this study, the development of the HOTS-oriented mathematics learning outcome assessment instrument is valid, reliable, and feasible to use.]]>
