We conducted a descriptive analysis of single-case research design (SCRD) studies on safety skills instruction (SSI) for individuals with autism spectrum disorder (ASD). Once we identified studies through electronic databases and reference lists, we used What Works Clearinghouse (WWC) Standards to evaluate each study. We analyzed studies in terms of various descriptive variables, calculated effect sizes through improvement rate difference (IRD), and aggregated effect sizes across studies to produce an omnibus effect size. Results showed 18 of 29 studies met the WWC Standards to "meet design standards" (MS) and "meet design standards with reservations" (MS-R), and various types of SSI were effective in teaching various skills. Of 18, 12 studies resulted in a large effect, and we found a behavioral skills training package (BST) to be evidence-based when we applied a 5-3-20 rule. Implications for researchers and practitioners are discussed.