This paper presents a new approach to behavioral choice modeling using arti
ficial neural networks (ANNs). A feed-forward network trained by a back-pro
pagation learning algorithm is used in this study. As a modeling technique,
ANNs are highly adaptive and very efficient in dealing with problems invol
ving complex interrelationships among many variables. The application of AN
Ns in the development of mode choice models is tested on the U.S. freight t
ransport market using information on individual shippers and individual shi
pments. Shipments are disaggregated at the 5-digit Standard Transportation
Commodity Code (STCC) level, representing the most detailed information pub
licly available. Results obtained from this exercise are compared with simi
lar results obtained from conventional legit and probit disaggregate mode c
hoice models. ANNs produced slightly better results compared with both legi
t and probit models. A method for analyzing ANN results based on examining
variable link weights is described. The method allows for increasing the ef
ficiency of ANNs by selecting only those input variables which significantl
y contribute to the network output. ANN mode choice models are expected to
behave equally well in the passenger transport market, both in the urban an
d intercity travel contexts.