<p>The lack of sufficient monetary returns and high market uncertainty limit the willingness of forest farmers to participate in forest carbon sequestration projects. Although some scholars have proposed methods for forest management decisions and valuation under stochastic prices, most were based on the geometric Brownian motion or mean-reversion process assumptions, which are insufficient and must be revised to reflect future price dynamics. This paper used 12 stochastic differential equations with different volatility and in categories of mean-reversion, constant drift, and random walk to test their predictive performance on timber and carbon prices in China. Then, a dynamic programming algorithm was developed to solve the generalized Faustmann model with a stochastic price to achieve more accurate calculations for forest optimal rotation and valuation. The results showed that, in the context of China’s timber and carbon markets, neither the classic geometric Brownian motion nor the mean-reversion process is best for predicting future price dynamics. Futher, they underestimate optimal rotation and land expectation value. The Monte Carlo simulation results indicated that the differences in land expectation value vary greatly among different stochastic price models, thus emphasizing the importance of selecting an appropriate price model.</p>