We use 1994-1998 data from the Russian Longitudinal Monitoring Survey (RLMS) to measure the incidence and determinants of several types of worker training and to estimate the effects of training on workers’ interindustry, interfirm, and occupational mobility, their labor force transitions, and their wage growth in Russia compared to the U.S. We hypothesize that the shock of economic liberalization in Russia may raise the benefits of training, particularly retraining for new jobs, but uncertainty concerning the revaluation of skills may raise the costs, with an overall ambiguous effect on the amount of training undertaken. The RLMS indicates a lower rate of formal training than studies have found for the U.S., suggesting that the second effect dominates. Previous schooling is estimated to affect the probability of training positively, but the relationship is much stronger for additional training in the same field than for retraining for new fields, consistent with the hypothesis that schooling and training are complementary but become more substitutable in a restructuring environment. Additional training in workers’ current fields is estimated to reduce mobility and earnings, suggesting inertial programs from the pre-transition era. Retraining in new fields increases all types of worker mobility and has higher returns than those typically observed for training in the U.S., but it also raises the variance of earnings and the probability of unemployment, consistent with a search view of such retraining. Given the large returns to retraining, the efforts of Russian workers to learn new skills may increase as uncertainty is resolved and restructuring proceeds.