In Indonesia, several training pieces have been conducted to develop HOTS-based science questions for elementary school teachers. However, few studies examine how elementary school teachers developed HOTS-based science questions. Hence, the current study is a case study that aims to analyze HOTS-based science questions developed by elementary school teachers in Indonesia. This study employed a descriptive-analytical method by analyzing 60 HOTS-based science questions developed by elementary school teachers. This study involved two raters from one of Indonesia's universities, who are experts in assessing learning for elementary school level to assess the questions. The aspects that are assessed are aspects of conformity with HOTS questions criteria, conformity with the criteria of thinking ability of elementary school students, and conformity with the quality of good questions. This study's findings show that many problems do not meet the first aspect, especially at non-algorithmic points, multiple ways, and multiple interpretations. Next, in the second aspect, most of the question does not use illustration to strengthen the developed question's information. The third aspect is a relatively fulfilled aspect by most developed questions. However, only a few developed questions are judged not to meet this aspect. Hopefully, this study's findings can be an evaluation material for policymakers in education to increase the training quality of HOTS-based science questions for elementary school level in Indonesia.