Background and aimsThe use of standard dynamic root architecture models to simulate root growth in soil containing macropores failed to reproduce experimentally observed root growth patterns. We thus developed a new, more mechanistic model approach for the simulation of root growth in structured soil. Methods: In our alternative modelling approach, we distinguish between, firstly, the driving force for root growth, which is determined by the orientation of the previous root segment and the influence of gravitropism and, secondly, soil mechanical resistance to root growth. The latter is expressed by its inverse, soil mechanical conductance, and treated similarly to hydraulic conductivity in Darcy’s law. At the presence of macropores, soil mechanical conductance is anisotropic, which leads to a difference between the direction of the driving force and the direction of the root tip movement. Results: The model was tested using data from the literature, at pot scale, at macropore scale, and in a series of simulations where sensitivity to gravity and macropore orientation was evaluated. Conclusions: Qualitative and quantitative comparisons between simulated and experimentally observed root systems showed good agreement, suggesting that the drawn analogy between soil water flow and root growth is a useful one.