This article presents a critical review of the use of cognitive ability testing for access to graduate and higher professional occupations to promote further debate and reflection in both the academic and practitioner community. The main contentions are that the practice of applying cognitive ability testing in these contexts has strong potential to both maintain and exacerbate social inequality in access to higher occupations and professions, and that validity evidence does not justify this to the extent that has previously been presumed. Five critical observations are examined, namely (1) evidence of adverse impact in test outcomes; (2) the tendency to position cognitive ability testing early in selection processes in high-volume recruitment; (3) recent evidence challenging the meta-analytic validity of cognitive ability tests; (4) weaknesses in historical primary validity studies; (5) conceptually flawed examination of differential validity evidence in the literature. Implications for practice are discussed, contrasting strategies that involve modifying selection systems that include cognitive testing, versus removing and replacing cognitive tests.