We report on continuing research on the British scientific elite, intended to illustrate a proposed new approach to elite studies, and based on prosopographical data on Fellows of the Royal Society born from 1900. We extend analyses previously reported of Fellows’ social origins and secondary schooling so as to take their university careers into account. The composite term ‘Oxbridge’ is called into question, as Cambridge appears historically to have been far more productive of members of the scientific elite than Oxford. However, Fellows from more advantaged class backgrounds do have a clearly higher probability than others of having attended Cambridge, Oxford or London, rather than universities outside of ‘the golden triangle’ – an outcome only partially mediated through private schooling. The ‘long arm’ of family of origin is thus apparent, although private schooling has been more important in helping Fellows from managerial rather than from professional families to gain entry to an elite university. Family influences on Fellows’ fields of research also remain, even though a further major factor is the universities they attended. A ‘royal road’ into the scientific elite, which Fellows from higher professional and managerial families have the highest probability of having followed, can be identified: that leading from private schooling to both undergraduate and postgraduate study at Cambridge. But the most common pathway, taken by 20% of all Fellows, is that leading from state schooling to undergraduate and graduate study at universities outside of the golden triangle. Fellows from higher professional, but not managerial, families show a distinctively high probability of having avoided this pathway; but it is that most common for Fellows of all less advantaged class origins. The case of the British scientific elite would suggest that detailed and disaggregated analyses of processes of elite formation can show these to be much more diverse than has often been supposed.