Recent reports suggest that one-third to one-half of TBI study populations are not included in outcome samples because of the inability to find them after leaving inpatient treatment. The results of a preliminary study conducted at Ohio State University suggested that systematic bias in longitudinal studies may result from subjects with premorbid substance use problems being lost to follow-up at one year. Population estimates for return to work or school may be over-estimated if those lost who have substance use problems resemble those followed. The current study expands on this initial examination of potential biases from those lost to follow-up by extending the research to multiple samples of those followed at one- and two-year follow-ups.