This study aims at examining the dynamics of production structures in Cameroonian agricultural sector under stochastic climate change. More specifically we develop and estimate a dynamic demand system of inputs and output by using a stochastic dynamic model approach and applying the hotelling’s lemma to the Hamilton-Jacobi-Bellman equations. The data are compiled from Food and Agricultural Organisation (FAO), Cameroonian Ministry of Agriculture and Rural Development (MARD) and the National Institute of Statistics of Cameroon (NSI). The model is estimated by non-linear three least squares method. We found that under climate change and market price change, food crop production, capital, labour, arable land and fertilizers are actually quasi-fixed variables and their adjustments are done progressively. Then, crops adjust each year by 56% towards their desired level in response to climate change. While the capital factor adjusts by 10.32% towards its equilibrium level in one year. Furthermore, the labour is adjusted by 68.52% each year. The fertilizer is adjusted at 44.11% per year. Finally, the arable land adjusts to 87.32% per year. To further reduce the climatic risks to which Cameroonian farmers are faced, public decision-makers can reduce directly and indirectly adjustment cost by funding research on adaptation crops and land use.