Objective: Extreme impairment on the Trail Making Test (TMT) has been reported to be associated with invalid performance in adults with traumatic brain injury and those involved in head injury litigation (Iverson et al., 2002). The present study aimed to investigate whether the TMT provides effective embedded validity indicators in a sample of clinically-referred adolescents. Method: Data were analyzed from 132 adolescents, aged 15 to 17, with a wide range of diagnoses. Raw time-to-completion scores on TMT-A and TMT-B, as well as the ratio comparing time to completion on TMT-B versus TMT-A (TMT B/A), were evaluated against established performance validity tests. Results: Adult cut-offs (≥62” on TMT-A, ≥200” on TMT-B, and ≤1.49 on TMT B/A) all produced excellent overall specificity (.94–1.0). Although sensitivity was variable for the TMT-A cut-off (.00–.50), the TMT-B and TMT B/A cut-offs produced consistently low sensitivity (.00-.13). Performance on TMT-A and TMT-B was correlated with intellectual functioning (rxy = −.30–.31; rxy = −.22–.26), while TMT B/A covaried with gender (rxy = .20). New alternative cut-offs were also explored to increase sensitivity. Conclusion: While adult TMT cut-offs may be used as embedded validity indicators with clinically-referred adolescents given their consistently high specificity, their low and variable sensitivity limits their ability to reliably detect invalid response sets.