Accurate, real-time state of charge (SoC) and state of health (SoH) estimation is essential for lithium-ion battery management systems to ensure safe and extended life of battery packs. For the large battery packs associated with battery electric locomotives and grid applications, computational efficiency is critical, especially for onboard implementation. This paper presents real-time SoC and batch least squares SoH and current sensor bias estimation using measured cell voltage and current from large battery packs. An online gradient-based SoH estimator, coupled with the online SoC estimator, provides real-time onboard health monitoring. The online and offline SoC-SoH algorithms are tested using data from a battery electric locomotive. The SoC-SoH estimation results show tightly clustered capacity, resistance, and current sensor bias estimates for an 11-cell module. The batch and online capacity estimates match to within 5% after the startup transients decay.