Childhood bullying is an important predictor of psychological and health outcomes in adulthood; however, validated retrospective measures of childhood bullying are lacking. This study investigates the psychometric properties of an adult retrospective version of the California Bullying Victimization Scale (CBVS). The CBVS self-report measure was developed for use with children and adolescents to assess the three definitional characteristics of bullying (aggression that is chronic, intentional, and involves an imbalance of power), without using the term “bullying.” In the current study, we evaluate patterns of retrospective reports of bullying victimization, and compare results to a common definition-first measure of bullying. Concurrent validity and 4-year stability are addressed. In the fall of 2012, entering first-year students at 4 universities in the United States (N = 1,209; 65.2% female) completed the California Bullying Victimization Scale–Retrospective (CBVS–R) as part of an online survey. In spring of 2016, participants at 2 universities who provided contact information (N = 175) completed a 4-year follow-up survey. Results support the validity of the CBVS-R as a retrospective self-report measure of bullying victimization experienced in childhood. In particular, the percent of respondents classified as being bullied (27.9%) and age- and gender-related patterns of victimization were consistent with known patterns of childhood bullying. In addition, respondents reporting childhood victimization indicated increased psychological distress in adulthood. However, stability of reports across a 4-year follow-up period were lower than expected (κ = .38). Implications for the use of retrospective reports of childhood bullying victimization are discussed.