Undergraduate internships have gained popularity among students, universities, governments and firms since the creation of the European Higher Education Area. However, empirical research on the relationship between internships and labour market performance of graduates is still scarce, particularly in Spain. This paper examines whether internships improve job attainment in the short run (first employment after graduation) and in the medium/long term (employment 4 years later). We use the first Spanish University Graduate Job Placement Survey (2014) to estimate linear probability models and probit models. A novel econometric technique is also implemented to evaluate the sensitivity of our findings to omitted variable bias. We disentangle the internship effect on (i) the time it takes to find the first job; (ii) the vertical, horizontal and skills matching with the first job; (iii) being employed in the medium/long term; (iv) the vertical and horizontal matching with the current employment; and (v) wage quintiles of the current job. Our results show that internship experience facilitates the university-to-work transition for Spanish graduates. Although the effects of internships on being employed do not vanish in the medium/long term, there is weak evidence of positive effects on matching four years after graduation and no effects are found on wages.