This study compared student performance on language-usage test items presented in two different formats: as discrete sentences and as items embedded in passages. American College Testing (ACT) Program's Assessment experimental units were constructed that presented 40 items in the two different formats. Results suggest item presentation may not affect item difficulty to any significant degree, even though some language-usage classifications may be particularly susceptible to the impact of format. Questions about content validity are also considered. Samples of the two formats, passage presentation and discrete-sentence presentation, are included in the appendices. (Author/JAZ)