Performance criteria are critical variables in determining the effectiveness, durability, and generalizability of behavioral interventions. This chapter reviews systematic analyses, empirical studies, and the author’s recent review to examine how performance-criterion components—levels and frequencies—affect response maintenance and generalization of trained behaviors. Evidence consistently shows that higher performance-criterion levels (≥90% for skill acquisition; ≤10% for behavior reduction) and moderate frequencies (2–3 consecutive sessions) produce stronger and more stable behavioral outcomes. Lower levels or minimal/excessive frequencies are associated with greater variability and weaker maintenance and generalization. Performance criteria interact with participant and task characteristics, such as age, diagnosis, and skill type, suggesting that criteria must be tailored to the individual learner and the specific context. Chi-squared and regression analyses support the predictive role of stricter and moderately applied performance criteria in sustaining and transferring behavior across contexts. Collecting maintenance and generalization data, along with transparent reporting of performance-criteria components, is essential to ensure that acquired skills are durable, generalized, and socially meaningful. These findings emphasize that performance criteria are one of the functional study variables central to optimizing the outcomes of behavioral interventions.