The assessment of sexual offenders consists of the systematic collection of clinically relevant information in order to detect clinical problems and to provide clear treatment targets. A key but neglected issue in the area of sexual offending concerns the role of individual case formulations in the determination of offenders' treatment needs. In this paper, we investigate the relative strengths and weaknesses of manual-based treatment (MBT) and formulation-based treatment (FBT) for sex offenders. On the one hand, FBT has the advantages of greater flexibility and a more individualistic focus, and arguably is better equipped to deal with more complex clinical presentations. On the other hand, MBT has the advantages of standardisation and less reliance on clinicians' (flawed) judgement, and may be a more efficient use of scarce resources. We conclude that clinicians should initially provide manual-based treatment rather than that based on individualised case formulations. However, we also suggest that there are at least four situations where FBT represents a valuable strategy, namely when confronted with particularly complex or unusual cases, when standardised treatment has failed, or when there are significant threats to the therapeutic relationship. Finally, we briefly discuss some possibilities for research, and caution that our mixed model is only proposed as a temporary solution and that ultimately any model concerning treatment selection needs to be empirically based and conceptually defensible.