A lively methodological debate together with organisational adjustments has dominated the international discussion on development evaluation in the last decade. Less discussed have been the evaluation processes, from procurement of consultants to the completion of evaluations, where many donors have used the same ‘system’ with only minor adjustments for many years. A key question is whether this ‘system’ is still ‘fit for purpose’ in view of the increasing complexity of many development interventions. This article presents three perspectives on development evaluation processes – the ‘independence perspective’, the ‘transaction cost perspective’ and the ‘political economy perspective’ – and argues that in particular the ‘political economy perspective’ constitutes a relevant perspective in analyses of evaluation processes. It is concluded that evaluation processes should be designed to reflect the specific characteristics of the individual evaluations, but also that more analysis and experimentation are required.