Study attrition is a problem in all community-based intervention studies using longitudinal research designs, but is compounded with hard to reach populations. High attrition poses threats to internal and external validity and may result in an inadequate sample size. The purpose of our study was to determine the characteristics associated with attrition. The study employed data from a cross-site evaluation of jail diversion programs. A self-report interview was conducted at baseline for 1,289 individuals. A 33% and 52% attrition rate was observed at the 6-month and 12-month follow-up interviews, respectively. The characteristics associated with loss to follow-up were male gender, part-time or full-time employment, drug offenses, jail days, baseline interview location, community supervision, and community geography. Knowing which individuals are more likely to attrit allows evaluators to develop targeted sampling strategies and participant engagement strategies.