This research paper presents an interesting finding on the assumption of an average data for MTBF and MTTR to predict the automotive assembly line throughput. A simulation modeling study is presented here for an automotive robotic welding assembly line. The assembly line layout in this study consists of 10 process stations involving material handling robots, welding robots, sealing, metal piercing, sheet metal welding, optical gauging, and unload stations.  There is no buffers in-between the process stations. The parts are transferred by robotic “hand-off” process. The line under study is a completely automated assembly line with no operator involvement except for maintenance. The aim of this paper is to develop a simulation model using the Arena software for the assembly line. Also, develop various scenario analyses to investigate the effects of using the average, minimum and maximum values for the Mean Time Between Failure (MTBF) and the Mean Time To Repair (MTTR) to predict the line throughput. The simulation model prediction results show that the average value MTBF/MTTR assumption actually produce poor results. The simulation model results show that the productivity could be improved by 13% using the actual maintenance data instead of using average assumption data.