Knowledge of the optimal time lags following particular call outcomes in a telephone survey (e.g., callbacks, no answer, busy, refusal) are important for maximizing the number of completed interviews in a given sample and utilizing resources with maximum efficiency. Previous research estimating time lags usually treats each attempt as completely independent or dependent only upon the immediately preceding calling event, failing to recognize a probability structure dependent on call history. This work extends research by Sangster and Meekins (2004) that uses the proportional hazards of contact and interview completion/refusal in models incorporating a variety of call history variables and mean lag between attempts. This research estimates optimal time lags by modeling the probability of contact or completion using a survival model approach. Estimates are obtained from the call histories and lag times of respondents to the Telephone Point-of-Purchase Survey, a nationally representative, list-assisted RDD survey conducted by the U.S. Census Bureau for the Bureau of Labor Statistics.