Purpose - The present research aimed to investigate the effects of attribution on expert clinical judgment in comparison to semi-experts and laypeople. Two research questions were addressed. First, would experts be less subject to attributional manipulations, in terms of their perceived ratings of dangerousness, than would semi-experts or laypeople? Second, would experts be less subject to attributional manipulations, in terms of their assessments of offender responsibility, than would semi-experts or laypeople? Method - A 3×3×2 mixed groups design was implemented. Participants read nine crime scenarios that had been internally or externally manipulated. For each scenario, participants were asked to rate offender dangerousness, offender responsibility, and the seriousness of the crime and to suggest a suitable sentence length. Targeted recruitment was employed, yielding 12 experts, 21 semi-experts, and 22 laypeople. Results - Offenders were considered to be more responsible for their actions and more dangerous to others in the internal manipulations than in the external ones across all crime types and by all levels of expertise. Findings indicate that semi-experts are less subject to the influence of attributional manipulations than both experts and laypeople. Marked similarities in the pattern of expert and lay person judgments can be observed from the present analyses. Conclusions - The current findings lend support to previous research in the area in that similarities between expert and lay person judgment were observed. However, through expanding and clarifying the levels of expertise investigated, the current findings highlight the need for greater research into the distinct ‘semi-expert’ group.