Many university graduates conduct internships before starting to work in a direct-hire job. I analyse the effects of internships on early labour market performance to evaluate whether they enhance or hinder the university-to-work transition. I use propensity score matching to identify graduates that resemble each other in important characteristics such as cognitive ability, and only differ with respect to the internship experience. This allows comparison between interns and non-interns in key dimensions of job market performance: monthly earnings, employment status, and job satisfaction. The results suggest that internships have detrimental effects across dimensions. Graduates with an internship experience are significantly less likely to be employed one year after graduation, and, if employed, earn significantly less than their non-intern peers. However, the negative effects are short-lived and vanish within five years. Due to this catching up, I can rule out that interns are a negative selection of all graduates. Instead it seems that the internship sends a negative signal to prospective employers and is thus causing the underperformance at job entry. The measured effects are less pronounced for female interns.