Over time, there has been an increase in the share of trip chaining – tours that have at least one intermediate stop between the main tour origin and destination. For trip-based models, this shows up as an increase in the share of trips that are non-home-based. For tour-based models, it shows up as an increase in the share of complex tours. Data is available to substantiate both trends. NHB trips are difficult to model accurately because of their varied nature and the fact that they aren’t anchored at the home. Almost all aggregate four-step models treat the legs of a complex tour as being independent of each other, both in space and time. With the increase in complex tours, the author believes that this may lead to less accurate travel models. The reasons for the increase in tour complexity are not well known. It may be that this trend is caused by increased congestion and/or increased travel costs, encouraging people to travel in a more efficient manner. This trend is a key motivation for the movement towards disaggregate tour-based models, such as activity-based models. Such models explicitly account for trip chaining, thus removing the NHB trip category entirely. This is likely to lead to more accurate travel estimates. The author provides some evidence that this has occurred, by comparing trip-based and tour-based models of person travel in Brunswick, GA and truck travel in Atlanta, GA.