This paper summarizes and evaluates the results of experiments which consisted of the elimination of bus fares in off-peak in the cities of Trenton, New Jersey and Denver, Colorado. Because the data in Denver had significant weaknesses, the specific figures cited herein are for Trenton (more accurately, Mercer County, New Jersey); the results of the two experiments were essentially identical, however, so the conclusions drawn are made in a more general context. The evaluation found that the net ridership increase during the demonstration was on the order of 15% (about 45% during the off-peak periods). This included the combined effects of an increase in trip frequency by prior users and an increase in the number of off-peak bus riders. Most new bus trips were diverted from other modes; very few were newly generated. The observed effects on the operator were minimal, with the exception of the associated revenue loss. The effects on bus drivers included schedule adherence problems because of increased passenger boardings and problems with on-board security because of increased rowdyism among younger passengers. The findings indicate that the fare-free (off-peak) transit experiments did not significantly achieve most pre-conceived goals.