Today’s automotive industries require quick ramp-up in order to shorten the time to market and fulfill greater demand for product variety. Since, dimensional tolerance problems are one of the main causes for delay during ramp-up in multi-station assembly systems; hence, rapid diagnosis of faults contributing to dimensional errors is of significant concern. This paper presents an agent-based simulation model (ABSM) which integrates model-based and data-based diagnostics for fault diagnosis and correction in multi-station assembly processes. The fault causing variation sources, their effects on dimensional accuracy, diagnosis from sensors data and corrective actions in multi-station assembly systems are simulated using proposed ABSM. A case study has been considered to illustrate the ABSM methodology for fault diagnosis in multi-station assembly systems.