The 19 sets of data came from 12 empirical reports representing 2,082 subjects, and the review covered studies available as of March 1991. The analysis demonstrated a significant overall difference between weapon-present and weapon-absent conditions, with weapon presence leading to reduced identification accuracy. The size of the effect, however, was small (.13) for the dependent measure of lineup identification and moderate (.55) for feature accuracy. Data supported the hypothesized weapon focus effect. Theoretical implications of the findings for crime scene investigation and complexity are discussed, as well as factors that appear to mediate and facilitate the weapon focus effect. Appendixes contain supplemental study data. 16 references