Purpose -- The aim of this paper is to develop a detailed descriptive model for kitting operations, allowing resources sizing and computation of systems' economic performances. Design/methodology/approach -- A mathematical model allows to size resources, given product characteristics and production mix, and determines total system costs by assessing relevant cost items including investment costs (vehicles, containers, storage racks), direct operating costs (transport and kitting workforce, vehicles energy consumption and maintenance, quality costs), indirect operating costs (space requirements, work in process (WIP) and safety stock holding costs, administration and control). Findings -- The choice of parts delivery supply to assembly lines requires a thorough economic comparison of alternatives. However, existing models are often simplistic and neglect many critical factors which affect the systems' performances. As a consequence, industries are unsure about which system is best for their environment. This model allows assessment of the cost and suitability of kitting in any specific industrial setting. Results of the model application are case-specific and cannot be generalized, but the major impact of labour and error correction cost has been highlighted. Research limitations/implications -- The model at present focusses on the in-house kitting systems based on travelling kits concept only. Although all quantitative cost drivers are included, some context-related qualitative decision factors are not yet included. The model assumes that the information about product structure and part requirements is known and that a preliminary design of the assembly system (i.e. line balancing) has been carried out. Practical implications -- Production managers are given a quantitative decision tool to properly assess the implementation of kitting policies at an early decision stage. This allows exploring the trade-offs between the alternatives and properly planning the adoption of kitting systems, as well as comparing kitting with alternative material supply methods. Originality/value -- With respect to previous simplified literature models, this new approach allows quantification of a number of additional factors which are critical for successful implementation of cost-effective kitting systems, including kitting errors. An exhaustive cost estimation of kitting systems in multiple, mixed-model assembly lines is thus permitted. Paper type Research paper [ABSTRACT FROM AUTHOR]