We construct a model of rent-maximizing behaviour by a single seller of timber in the absence of a formal market, deriving the stochastic implications of rent maximization for timber prices (stumpage rates) when other input and output (lumber) prices are random. Subsequently, we examine the model's ability to describe monthly, time-series, stumpage-rate data from British Columbia, Canada between January 1979 and October 1999. Deviations of stumpage rates from their long-run trend are also structured by an error-correction model which suggests that between 13 and 20% of period-to-period changes in stumpage rates can be explained by an equilibrium adjustment term.