Self-administered questionnaires were used to obtain baseline data on substance use and sociodemographic factors from 2,838 men and women who were attending Navy basic training or Officer Candidate School in 1998. Follow-up data were acquired through questionnaires mailed in 2000. Up to three attempts per participant were made to elicit a complete survey. Areas covered in the questionnaire included background questions on sociodemographic factors, substance use (tobacco, alcohol, and drugs), normative beliefs about drinking, and occupational factors (stress, problems at work, and deployment). The dependent variable was panel attrition. Participants at baseline were categorized as "attritors" if they refused follow-up participation; failed to respond to numerous follow-up attempts; were not able to be contacted due to military transfer, bad address, etc.; or had separated from the Navy, in which case no further contact information was available. Findings showed that tobacco use was a significant predictor of attrition. A significant interaction between educational level and drug use indicated that respondents with less than a college education who were also drug users were at increased risk for attrition. Other significant predictors of panel attrition were being male and younger age. Alcohol use was not significantly associated with attrition. Thus, tobacco users and drug users with less than a college education may become a significant source of nonresponse-bias in longitudinal surveys of employed young adults. 2 tables and 41 references