Background: Understanding of informed consent forms (ICFs) for clinical research remains insufficient despite attempts to simplify them. Through linguistic discourse analysis, we sought to identify pitfalls within the text of ICFs that could hinder readers' understanding of participation in research. Methods: We conducted a linguistic discourse analysis on a qualitative sample of 19 ICFs approved by research ethics committees (RECs) for oncology protocols and explored whether our findings also applied to standard U.S. documents available online. Results: We identified five major categories of language patterns that were problematic with respect to ensuring informed consent. We categorized them as follows: “bypassing consent,” “seeker–supplier inversion,” “interlocking Russian dolls,” “vanishing author,” and “one size fits all.” At least one instance of these findings existed in all analyzed forms (median 10 per ICF, range 1–18) and in national templates and U.S. documents. Conclusions: Linguistic discourse analysis identified recurrent pitfalls in the language of REC-approved ICFs and templates. This approach may provide new tools to improve ICFs.