This paper develops a random utility based estimation framework for the Household Activity Pattern Problem (HAPP). Based on the realization that output of complex activity-travel decisions form a continuous pattern in space-time dimension, the estimation framework is treated as a pattern selection problem.