A virtual version of the Trier Social Stress Test (TSST) has been increasingly used in stress research. Benefits of the virtual TSST include that it is more economical, and offers improved control of context and enhanced flexibility of use. Many studies have confirmed the applicability of virtual TSSTs in stress research, but only in adulthood. In the present study, we aimed to experimentally verify the transferability of a virtual TSST to adolescence. A total of 73 healthy adolescents (aged 11–17 years) of both sexes completed either a real (IV-TSST-C) or virtual (VR-TSST-C) TSST for children. The surveyed stress parameters included salivary cortisol and alpha amylase concentrations, heart rate, heart rate variability, and subjective stress ratings across test sessions, as well as a pre- to post-TSST-C comparison. All parameters revealed significant stress responses over time. We observed significant effects of group on cortisol and subjective stress ratings, with the VR-TSST-C inducing less stress than the IV-TSST-C. In alpha amylase, heart rate, and heart rate variability, we found no significant difference between TSST versions, but only significant equivalence for alpha amylase. Some parameters may have been influenced by an expectancy effect and the participant’s sex. In conclusion, among adolescents, a virtual TSST-C effectively elicits stress at multiple levels (endocrinological, autonomic, and subjective); however, the magnitude is not always comparable to with the real TSST-C, which should be considered when applied.