The modelling of marine zooplankton has made great progress over the two last decades covering a large range of representations from detailed individual processes to functional groups. A new challenge is to dynamically represent zooplankton within marine food webs coupling lower trophic levels to fish and to thereby further our understanding of the role of zooplankton in global change. In this respect, the “rhomboid strategy” (deYoung et al., 2004) has been suggested as a generic approach to model the various trophic levels of pelagic ecosystems and is deemed to be adaptable to different spatial and temporal frames of applications. The present paper identifies directions to develop zooplankton modelling by combining the skills of modellers, experimentalists, observers and theoreticians. In the first part, we present the main types of existing models, specifying the scientific issues, their characteristic time and space scales, across the ecological organization levels. In the second part, we focus on the strengths and weaknesses of parameterizations for the different processes. Finally in the last part, we make suggestions for improving these parameterizations by combining experiments and observations, using modelling techniques to transfer information across scales and testing theories which can themselves help to organize experimental and modelling research.