Locomotive prime mover engine emission rates are typically measured at steady-state for discrete throttle notches using an engine dynamometer weighted by a standard duty cycle. However, this method may not represent real-world locomotive emissions. A method for in-use measurement of passenger locomotives, using a portable emissions measurement system (PEMS), was developed to estimate duty cycle average emission rates. We conducted 48 measurements of one-way trips between Raleigh and Charlotte, NC, on 7 locomotives and 18 sets of measurements in the rail yard (RY). Real-world duty cycles differed from those used for regulatory analyses, leading to statistically significant lower cycle average NOx and HC emission rates. Compared to RY measurements, notch average NOx emission rates measured over-the-rail (OTR) at the highest two notch settings were, on average, 19% lower for four locomotives. At the highest notch, OTR CO2 emission rates were, on average, 12% lower than RY rates for five locomotives. For a more accurate representation of real-world emission rates, OTR measurements are preferred. However, using steady-state notch average RY emission rates and standard duty cycles may be tolerable for some applications. OTR versus RY cycle average emission rates typically differed by less than 10%.