To obtain a better understanding of the factors affecting asparagus spear growth and yield, a process-oriented and stochastic model for asparagus ( Asparagus officinalis L.) grown under soil ridges was developed and tested. This model describes a population of asparagus plants with a binomial distributed bud cluster number per plant. Each bud cluster starts to grow after exposure to an exponential distributed temperature sum. The length growth response of spears, which is mainly dependent on soil temperature and spear length, is affected by the soluble carbohydrate ( CHO) concentration in the storage root system and the temperature gradient between the spear tip and the rhizome. The spear growth rate is assumed to be normally distributed, while a stochastic dependency to the required temperature sum for bud break is assumed. For each bud cluster, a broadly varied apical dominance relation between successive spears is described by an exponential distributed temperature sum for spear length growth induction. After removal of the dominating spear, the required temperature sum decreases exponentially. The time progression of spear diameter is a function of spear generation number within a bud cluster and the mean temperature during spear growth induction and can be used to derive the spear fresh weight. The change of soluble CHO concentrations in the storage root system is derived from maintenance and growth respiration requirements. Some of the model parameters are derived from independent measurements of spear growth. Regarding the tremendous variability of daily yields, the model performed well in the prediction of daily spear number, diameter and yield over the whole harvest season across two production sites with various soil temperatures induced by different types of ridge covering. The effects of soluble CHO concentration and soil temperature gradient on spear growth rate are regarded as preliminary and should be tested in further experiments. Due to the apical dominance relation between spears, asparagus is subject to endogenous rhythmic growth, which invalidates simple statistical yield forecast approaches. In its current state, our yield forecast model is well suited to test the impact of various plant traits on asparagus yield, and thus to guide early selection in crop improvement projects. Finally, the current model could be used to derive easier to handle yet physiological sound yield models to be useful for field-scale applications.