Information on 288 participants who had been retrained between November 62 and April 1963, was collected by personal interviews and mailed questionnaires, and by earnings histories provided by the Social Security Administration, to discover the actual trends in average retraining benefits and to determine if this trend differed on characteristics of trainees. A basic regression was used to explain earnings in each of the years from 1962-66, considering such variables as sex, marital status, job status of nontrainees, age, education, number of dependents, extrapolated 1960 earnings, length of training course, and training status. It was found that benefits from retraining continued to accrue for each of the five years following the course, with an average gain in earnings of $5,834. The average gain from retraining for the five year period was only influenced by age, with greatest increases in earnings among trainees between 30 and 35. This influence of age did not become significant until the third year following the courses, illustrating the limitations of benefit-cost analyses which focus on short term gains. Recommendations are made upon these findings. (pt)