This paper uses cointegrated time-series methods to evaluate the effect of timber employment on participation in a major poverty program-Aid to Families with Dependent Children-Unemployed Parent (AFDC-UP). The study is conducted for major timber-producing counties in California. It is shown that a two-sector structural model can be solved to produce an error-correction model. An error-correction model is estimated with time series on state and county AFDC-UP caseload, state employment, county nontimber employment, and county timber employment. Utilizing tests on the cointegrating space, it is shown that there is no long-run relationship between poverty and timber employment in 10 of the 11 counties studied.