We present evidence from a six-year field experiment with young job seekers in urban labor markets in Uganda. We study how standard labor market interventions impact their search behavior, long run labor market outcomes, and how these long run outcomes are mediated through search behavior. The interventions we consider are the offer of vocational training, vocational training combined with job assistance, and job assistance only. Training is offered in sectors with high wage firms. Job assistance comprises a light touch offer to match workers for job interviews with such high wage firms. At baseline, youth are unskilled yet overly optimistic about their job prospects. Relative to controls, those offered vocational training become even more optimistic, search more intensively and direct their search towards higher quality firms. However, for youth additionally offered job assistance, expectations and search effort are revised downwards as call back rates from firms the workers are matched to, are far lower than their prior expectation. These differential search strategies impact long run outcomes: vocational trainees without job assistance have higher employment rates, longer employment spells, and end up in higher quality jobs and firms than workers additionally offered job assistance. Our analysis highlights how the potential for labor market entrants to be exuberant or discouraged both matter for long run outcomes through job search behavior. We discuss implications for the design and targeting of labor market interventions meant to help young people find good jobs.