This research provides insight into one US state's effort to incorporate higher-order thinking on its Algebra I End-of-Course tests. To facilitate the inclusion of higher-order thinking, the state used "Dimensions of Thinking" (Marzano et al., 1988) and "Bloom's Taxonomy" (Bloom et al., 1956). An analysis of Algebra I test items found that the state's initial interpretation and application of "Dimensions of Thinking" and "Bloom's Taxonomy" was faulty and inconsistent; as a result, few Algebra I test items from 1998 and 2001 were found to assess higher-order thinking. Algebra I test items written in 2007 were found to be more cognitively complex. This case study provides several findings that other test item writers (whether for the classroom or large-scale assessments) can apply to their own situation regarding assessing for higher-order thinking. These includes the importance of: (a) considering the "familiarity" a student has with a test item in deciding whether the item is lower- or higher-order thinking; (b) using a mathematics specific assessment framework with a limited number of categories; and (c) not confusing test items placed in a "real-world" context with higher-order thinking. (Contains 5 tables and 1 figure.)