This study evaluates the consequences of financial variables on the efficiency of Class I railroads in the United States for the period 1996-2006. A panel stochastic frontier analysis is used to simultaneously estimate the stochastic frontier model and financial ratio model with output and efficiency measures as endogenous variables. Results show the average efficiency measures was 83 percent across six major class I railroads. The Burlington Northern-Santa Fe was most efficient and Norfolk Southern the least efficient for the period, 1996-2006.