This study aimed to: (1) produce higher order thinking skill (HOTS) assessment instruments in learning Indonesian history; (2) know the validity of HOTS assessment instruments in learning Indonesian history; and (3) find out the characteristics of HOTS questions in learning Indonesian history. This study employed the research and development method of the Borg and Gall model. The HOTS test item was conducted on 36 students in class XI of 2 Ngaglik State Senior High School. Data analysis includes tests of validity, reliability, level of difficulty, distinguishing features and deception index. The study found: (1) the HOTS assessment instrument of multiple-choice questions consisted of 25 items; (2) the results of the HOTS question validation by two Indonesian history learning assessment experts on the material, construction and language aspects were valid and appropriate. The results of the validation by three Indonesian history teachers also stated that the assessment instruments were valid and appropriate; and (3) the characteristics of HOTS questions had fulfilled the validity criteria of 23 questions, reliability with a coefficient of 0.97 (very strong), the average difficulty level is 0.33 (moderate), the average differentiation test is 0.42 (good), and the average deception index is 0.56 (good).