Supervision of sexual offenders can only work to reduce risk when it monitors and addresses factors related to both general and sexual recidivism. It is well known that many sexual offenders commit other types of offences, such as violent and general offending, but other types of offenders rarely commit sexual offences (Hanson & Bussiere, 1998). It is therefore necessary to distinguish sexual offenders from other offenders when we study the different recidivism types, and the key risk factors for the prediction of any reoffending. This study assessed the predictive utility of several commonly used psychometrics in Northern Ireland, namely the Stable and Acute 2007, Risk Matrix 2000, and the STEP battery. Risk assessments were collected from a sample of 325 participants each of whom had been convicted of a sexual offence in Northern Ireland. The data is archival, sourced from risk assessments and psychometrics conducted on offenders subject to supervision under the Public Protection Arrangements for Northern Ireland (PPANI) between 2008 and 2010. Overall levels of risk and individual risk factors as measured by these instruments were compared to rates of reoffending. A number of salient individual factors were identified from the sample, such as capacity for relationship stability, sexual deviancy, rejection of supervision and victim access, which links to distinguishing typologies of offending in sexual offenders supervised within the Public Protection Arrangements for N. Ireland (PPANI). While it was not possible to statistically link individual factors 9 to re-offending rates, results indicated that overall risk levels obtained by all three assessment tools have predictive utility in relation to non-sexual offending and breaches of probation conditions or licensing. Statistical analysis of sexual re-offending was not possible due to the small number of such offences within this sample. Findings from this study have both strategic and practical implications for the management of sexual offenders in N. Ireland. The main findings were that all three instruments predicted offending behaviour, breaches of probation, or breaches of licence. Recommendations and directions to inform future policy and practice are outlined in the Discussion Section.