This article describes the process of developing and testing the Poverty Attribution Survey (PAS), a measure of poverty attributions. The PAS is theory based and includes original items as well as items from previously tested poverty attribution instruments. The PAS was electronically administered to a sample of state-licensed professional social workers. The three scales of the PAS--individual, cultural, and structural--demonstrated strong internal consistency reliability. We conducted exploratory factor analyses and forced three-factor analyses and retained items with factor loadings at or above 0.50. The items on each scale loaded as predicted with few exceptions. The PAS is a new tool for standardizing poverty attribution research. We discuss implications for social work education, research, and practice.