In healthcare priority settings, early access to treatment before reimbursement decisions gives rise to problems of whether negative decisions for cost-effectiveness reasons should result in withdrawing treatment, already accessed by patients. Among professionals there seems to be a strong attitude to distinguish between withdrawing and withholding treatment, viewing the former as ethically worse. In this article the distinction between withdrawing and withholding treatment for reasons of cost effectiveness is explored by analysing the doing/allowing distinction, different theories of justice, consequentialist and virtue perspectives. The authors do not find any strong reasons for an intrinsic difference, but do find some reasons for a consequentialist difference, given present attitudes. However, overall, such a difference does not, all things considered, provide a convincing reason against withdrawal, given the greater consequentialist gain of using cost-effective treatment. As a result, patients should be properly informed when given early access to treatment, that such treatment can be later withdrawn following a negative reimbursement decision.