Aim. The present study aimed to explore the factorial structure, validity and stability of the Hospital Anxiety and Depression Scale (HADS) in chronically ill and healthy Polish adolescents. Methods. 146 chronically ill (girls: 57.6%) and 309 healthy (girls: 45.9%) adolescents aged between 12 and 16 years (M = 14.03; SD = 1.3) filled in an adapted version of HADS (HADS-Teen) in hospital or school settings. The one-week test-retest reliability and construct validity was analyzed in two sub-samples of healthy adolescents. Results. HADS-Teen showed a two-factor structure in the chronically ill sample and a three-factor structure in the healthy sample. The Anxiety scale had high internal reliability and stability and adequate correlation with another measure for generalized anxiety. Still, the Depression scale had good stability, but poor internal reliability in both samples. In the healthy sample, the Depression items split into two factors: depressed mood together with psychomotor retardation/agitation and anhedonia. Conclusions. The issues concerned with the factorial structure of HADS are replicated in Polish adolescents as well. HADS-Teen shows different structures in chronically ill versus healthy adolescents. Results indicate that a special attention must be paid when assessing depression symptoms in healthy adolescents using this instrument.