Driverless transit features vehicles or transit units functioning with no onboard intervention from a driver or attendant. In comparison with traditional transit, driverless transit offers reduced labor costs and smaller vehicles, albeit with higher capital costs, and better performance records. As of 2005 there were 37 driverless transit systems in urban service, with five in the United States. Fifteen cities in Canada, Europe, and East Asia had driverless metros. To move driverless transit into the mainstream, research is needed in multiple areas. This paper examines research needs identified in 12 papers presented at the 10th International Conference on Automated People Movers (APMs), with supplemental information from related literature. Integrating APMs into airport security infrastructure, the development of lightweight materials for vehicles and guideways, and personal rapid transit (PRT) reliability and systems theory are among the high-priority research challenges. Medium-priority research needs would focus on special applications of and revenue generators for driverless transit. Although a great deal of driverless transit research has been conducted in simulation, there is a need for operating and performance data from existing systems. These data would enable driverless transit to be considered in modal alternatives analysis. Also, to bring concepts such as PRT into reality, full-scale models need to be tested in operational environments. It is argued that research supported by industry and local agency consortiums can make incremental advancements, but extensive studies require committed, high-level government funding. While the development of driverless transit in the United States has been stifled by high-profile cost overruns and abandonments, successful applications in Europe or elsewhere may revive domestic investment. Potential driverless transit research funding resources are suggested.