Inter-rater reliability (IRR) of the Structured Assessment of Violence Risk in Youth (SAVRY) amongst mental health professionals was explored. Participants (n=22) rated vignettes varying in violence severity and gender (case characteristics). Professional characteristics including perception of their confidence and objectivity in ratings were measured. Using ICC1 scores, the IRR was low for items when interpreted using the 95% confidence interval but much higher using individual ICC scores. Case characteristics did not conclusively affect ratings. Self-reported confidence in ratings was not associated with increased reliability. This suggests that professionals completing risk assessments are likely to benefit from ongoing training.