We develop a revealed preference approach to elicit science and engineering PhDs’ preferences over employment outcomes, exploiting cohort size variations. Depending on whether pecuniary and non-pecuniary rewards are sticky or not, increments in the PhDs’ cohort size decrease either the availability of their ideal employment categories or the related compensations. In both cases, the PhDs’ preferred employment categories are revealed to be the ones that are relatively less chosen when the PhDs’ cohort is large and relatively more so when it is small. Examining two major European universities, we find that PhDs equally value employment in highly-ranked universities and R&D-intensive companies. Moreover, these employment categories are preferred to low-ranked universities, non-R&D-intensive firms, and public administration. There is preference heterogeneity across PhDs depending on their research field.