In recent years, an increasing portion of bank income has been generat
ed through nontraditional activities. Most studies of bank efficiency
do not include any measure of nontraditional activities when measuring
outputs. In this paper, cost, revenue, and profit efficiency are esti
mated by using models with and without nontraditional output. The resu
lts suggest that the standard model which omits nontraditional output
understates bank efficiency. Evidence also surfaces that based on effi
ciency, the relative ranking of individual banks changes when nontradi
tional activities are included as a type of output. (C) 1998 Elsevier
Science B,V. All rights reserved.