Our study of cognitive aging involves a comprehensive battery of neuropsychological measures. The Paced Auditory Serial Addition Test (PASAT) is the only test that some of our participants will refuse to complete. We explored variables related to quitting versus completing the PASAT in this sample of normal older adults. We hypothesized that quitting would be related to personality features, subclinical anxiety symptoms, demographics, and/or Wechsler Adult Intelligence Scale-Revised (WAIS-R) Arithmetic performance. A logistical regression model including NEO Personality Inventory-Revised (NEO-PI-R) depression and excitement seeking, age, and WAIS-R Arithmetic classified participants with moderate accuracy. We encourage investigators involved in longitudinal studies to consider causes for missing data, especially when secondary to participant refusal.