The ability to estimate travel time on-line for use in signal timing optimization and route guidance and vehicle navigation applications is becoming necessary. Detector data are considered a valuable source of information on traffic conditions in transportation facilities. The development of models that use detector information to estimate travel time in urban networks is traced. Existing research efforts are briefly described and interrelated. A comparative discussion of the alternatives is aimed at providing a qualitative evaluation of a range of options available today for developing arterial travel time functions. The need for further calibration and validation of existing models is identified, and enhancements to improve their quality and applicability are recommended.