It would be vainglory to suggest that recent criticism has succeeded in justifying the large claims it has made for Shakespeare’s Romances. Though we may be convinced, because of the constant insistence, that the Romances are important, it is hard to point to the critic who has shown where the importance lies. At any rate, a retrospect of this century’s work on the last plays has little progress to report. “We cannot enlarge our conceptions,” said Hazlitt, “we can only shift our point of view.” The chronological story of the changing attitude to the Romances is often told; to repeat it here would give a false impression of organic development. The views of Dowden, Strachey and Wilson Knight do not act out Tillyard’s tragic scheme of Prosperity, Destruction and Regeneration. Although critics may see themselves as moving on from positions already reached, confounding the errors and enlarging on the hints of those who have preceded them, the chart of criticism over the years shows no continuous course, but a series of different vectors from different starting points. To make any sense out of the record of twentieth-century criticism of Shakespeare’s last plays, with its bewildering disagreements on what the plays contain, it is essential to discuss criticism in the light of the assumptions the writers make about literature. This essay is more of an attempt to distinguish between the prevailing critical attitudes to the Romances in the last fifty years than an account of work done. I have chosen some two dozen studies in order to illustrate the four or five main critical approaches and the different conclusions which writers using a similar approach may come to. I am particularly sorry that this scheme, partly because it is rather arbitrarily selective and partly because it gives weight to criticism relating to the Romances as a group, does not allow me to discuss many important studies of individual plays.