Current methodologies for timing signal control and for capacity analysis of signalized intersections rely on various formula-based delay models. Average stopped delay and average approach delay are two often used measures of effectiveness in such models. A common practice in estimating average approach delay is to assume first that vehicles arrive randomly at an intersection controlled with a pretimed signal. The estimates based on this assumption are then adjusted according to actual traffic pattern and type of signal control. The 1994 U.S. Highway Capacity Manual converts estimates of average approach delays into average stopped delays by using a conversion factor of 1.3. This practice may lead to large estimation errors. To add to the confusion about vehicle delay, the 1997 U.S. Highway Capacity Manual, which is being prepared, will use control delay as the measure of effectiveness. The relationships among these three types of delay are examined.