Noise pollution resulting from technological development, urbanization and economic growth is one of the major sources of complaints in urban areas. The sound generated by transportation systems, is one of the most important causes of noise-induced annoyance, since the exposure to high levels for long periods of time can be detrimental to health. Freight railway systems have great potential of noise emission, since they are designed to meet safety requirements, rather than comfort, and are subject to more severe operations and cruder maintenance procedures than passenger cars. Among the different types of noise that originate from a railroad, the squealing generated in curves is one that stands out since it can exceed regular rolling noise in 30 dB and often occurs in frequencies where the human hearing is more sensitive. Analytical models have been developed over the years to help understanding and predicting squeal. This paper aims to validate, at a freight railroad, an analytical model created to predict squeal sound pressure level (SPL) from subways and urban trains. The model was implemented on numerical computation software, taking into account wheel and bogie characteristics, train speed and curve radius. The estimated data were compared with measurements of SPL carried out at the railway during regular traffic. The squeal was heard during the measurement, and the noise peak was later identified during the data analysis. Measurements were carried out at three curves with different radii to investigate if the model accuracy was influenced by this variable. Nonetheless, since the occurrence of squeal noise results from the interplay of many parameters, it was not heard in some of the measurements. Even though, the results suggest that the model may be a useful tool to predict and understand the relationship between train speed, curve radius and squeal SPL.