Teachers need technological pedagogical content knowledge (TPACK) for teaching with technology, and its assessment is crucial for research and practice. Previous literature reviews on TPACK assessment were not specific to a content area (e.g., mathematics), although, by definition, the TPACK framework includes content-specific knowledge facets. Consequently, requirements for TPACK assessment could differ depending on the content. Further, reliable assessment of mathematics-specific TPACK depends on the quality of the test instruments used, but there is no consensus on the type of instruments used in past studies. This systematic literature review adds to existing reviews by focusing on TPACK assessment in mathematics, investigating the "study characteristics," the "instrument characteristics," and the "operationalizations" of TPACK. Regarding study characteristics, the findings reveal an increase in the number of studies conducted across various countries worldwide. As for instrument characteristics, researchers frequently used self-developed instruments to assess TPACK, often without providing information on reliability or validity measures. Regarding operationalizations, more than half of the studies used self-report scales to assess mathematics-specific TPACK, followed by observations and material analyses, while knowledge tests were hardly used. Additionally, assessments typically referred to the domain of mathematics as a whole instead of subdomains of mathematics. The results raise questions regarding the validity of the instruments and the comparability across studies.