181 participants aged 13–16 years old were recruited from a Scottish residential centre for young males with severe emotional and behavioural difficulties. They completed a health assessment and the Multidimensional Anxiety Scale for Children (MASC). Three observations over 24 weeks were conducted to evaluate the effectiveness of the MASC self-rating scale as a screening instrument for vulnerable adolescent males. Over the three observations, further comparison of the MASC detection rates of anxiety were made with the findings of the health assessment. The MASC failed to identify any of the young people who had a clinical diagnosis of anxiety previously assessed by the child and adolescent mental health services. Notably the MASC only identified 5–9% of the young people with current concerns raised about anxiety. This raises issues about the sensitivity of the MASC in identifying anxiety in this vulnerable population.