One size does not fit all in assessment and intervention for people with convictions for sexual offences. Crime scene indicators and risk-related variables have been used to identify distinct clusters of people with convictions for sexual offences, but there is a need for more robust typologies that identify clusters based on psychologically meaningful risk factors that can be targeted in treatment.

To use robust modelling techniques to identify latent profiles of people with convictions for sexual offences based on indicators of dynamic risk.

Adult male participants, who had been convicted for sexual offences and assessed for eligibility for the prison-based Core Sex Offender Treatment Programme delivered by His Majesty's Prison and Probation Service (UK), were randomly allocated to a test (n = 1577: 70.2%) or validation (n = 668: 29.8%) data-set. Exploratory factor analysis (EFA) was used to select measures of dynamic risk from psychological test data. EFA indicated four factors, from which six measures were selected for inclusion in latent profile analysis.

Five latent profiles were identified in the test and validation data-sets. These were labelled low psychological impairment, impulsive, distorted thinker, rape preoccupied and child fantasist. Profiles varied in individual characteristics, offence histories, victim preferences and level of risk.

Our findings should be used to guide assessment and intervention practices that are tailored to distinct psychological profiles consistent with principles of risk, need and responsivity.

对性犯罪定罪者的评估和干预并不能一刀切。犯罪现场指标和风险相关变量已被用来识别有性犯罪定罪的不同人群，但需要更强大的类型学来根据可作为治疗目标的具有心理意义的风险因素来识别人群。

使用强大的建模技术，根据动态风险指标来识别性犯罪定罪者的潜在概况。

曾因性犯罪被定罪并接受英国国王陛下监狱和缓刑服务局提供的监狱核心性犯罪者治疗计划资格评估的成年男性参与者被随机分配到测试 (n = 1577: 70.2%) 或验证 (n = 668：29.8%）数据集。探索性因素分析（EFA）用于从心理测试数据中选择动态风险的度量。 EFA指出了四个因素，从中选择了六项措施纳入潜在概况分析。

在测试和验证数据集中确定了五个潜在特征。这些人被标记为低心理障碍、冲动、扭曲的思想家、沉迷于强奸和儿童幻想家。个人特征、犯罪历史、受害者偏好和风险水平各不相同。

我们的研究结果应用于指导评估和干预实践，这些实践是针对符合风险、需求和响应原则的不同心理特征而定制的。