Objectives: Screening women for Chlamydia trachomatis (CT) infection using selective screening criteria has been operational in the northwestern United States (Region X) since 1988. Changes in the field, including declines in CT prevalence, introduction of sensitive laboratory tests, and budgetary pressures necessitate reevaluating the selective screening approach to ensure program credibility and efficiency. Goals: The goals of this study were to assess 1) performance of screening criteria in Region X, 2) predictors of CT infection, and 3) optimization of these criteria. Study Design: We conducted cross-sectional analysis of 409,882 CT test records of women from 1998 to 2000 using multivariate logistic regression and sensitivity and efficiency analyses. Results: Young age (<25 yrs), cervical signs of infection, and recent exposure to or history of chlamydial infection were strongly associated with testing positive. Behavioral risks showed a weak association with infection. Currently used selective screening criteria were sensitive but not efficient. Criteria weighted toward young age, exposure to chlamydia, or cervicitis would increase criteria efficiency by nearly 25% in some settings while detecting >90% of infections. Conclusion: Evaluating selective screening criteria can result in modifications that could increase screening efficiency.