To date, nematode dynamic models have been very simple, driven only by few parameters without accounting for host quality or environment characteristics. However, these approaches provided only a basic description of'nematode population dynamics, and a few mechanistic insights into the relation between the nematode, host and the environment. Recently, more specific models were developed for a wide range ofplant-parasitic nematodes and horticultural systems (Pratylenchus penetrans in rotations; potato cyst nematode and Meloidogyne incognita on potato systems; Radopholus similis, Pratylenchus coffeae and Helicotylenchus multicinctus on banana based systems). These models are either based on biological processes (population growth, initial or maximal population) or statistical approaches; often with dynamic outputs. These models account fqr the specificities of the relation in the nematode-plant complex, e.g. through the root biomass fluctuation which represents the food resource for nematodes. After reviewing the existing models that simulates nematode dynamics in hOliicultural systems, we present the example of the SIMBA-NEM model dedicated to plant-parasitic nematodes in banana based systems. We highlight the way this modelling approach allows integration of existing knowledge and permits us to re-examine research about nematode-plant relationship. We emphasize how these models may help to optimize the effect ofnematicide applications and participate to the design of sustainable and more environmental-friendly cropping systems. We also focus on the use of models to tackle issues surrounding new banana varieties. Finally, we discuss the relevance of the modelling scale from the root to the field, and its implication in the efficiency in forecasting population dynamics and plant damages. We examine the .needs for spatially explicit models that take into consideration the spatial variability of soil moisture or the availability of host root biomass. (Texte intégral)