A methodology for determining the effect of one term of military service on men's civilian earnings and occupations. Determining civilian benefits from military training would enable the Department of Defense to structure enlistment and reenlistment bonuses more efficiently, and provide background for veterans policy. Preliminary results show that military training has little effect on overall earnings, unless controlled for occupation. The proposed analytic method determines effects specific to civilian occupations. Results of an annual survey of first-termers who left the service a year earlier are matched against the respondents' service records to determine individual attributes. Results include (1) regression of earnings on personal attributes; (2) comparison of two electronics subfields; and (3) a partial tabulation of civilian occupational distributions for different military specialties. Unlike the usual earnings analysis, this approach can handle the various kinds of employment/nonemployment: occupation; job search; technical and formal schooling; reenlistment; and nonparticipation in the labor force. 29 pp. Ref.