This article attempts to describe the state of the art in building efficient dynamical plant growth and architecture models that contain the basic knowledge coming from botany, ecophysiology, agronomy, applied mathematics, and computer sciences. Plant architecture is the result of two processes working together: plant development, that concerns meristem functioning and speed of organogenesis which primarily depends on temperature, and plant growth, that concerns increase in biomass, driven by photosynthesis of the leaves, which primarily depends on light. The laws of plant development have been studied in detail by botanists through the resulting plant architectures; meanwhile, from their side, ecophysiologists have studied the role of the main physical factors (light, temperature, water) that control the yield of a crop. Thanks to the necessary simplifications required to build efficient mathematical models, botanists and ecophysiologists have merged their knowledge to generate truly functional and structural models of plants that allow to compute both their three-dimensional architecture and their biomass production in a variable environment. Applications for these crop models are numerous, from genetics to agronomy, but their usefulness depends mainly on the possibility to inverse the model for parameter optimization. Overcoming this bottleneck leads to the characterization of a plant with a set of functioning parameters that can be used for improving various issues involving selection, yield optimization, and optimal control for production in glasshouses or open fields.