Measures of adolescents' HIV/AIDS knowledge that cover only die basic facts may fail to assess underlying misunderstandings in need of intervention. We developed and evaluated a measure covering the broad domain of HIV/AIDS knowledge, emphasizing misunderstandings revealed in semi-structured qualitative interviews. These included the cumulative and relative risks of unsafe behaviors, proper condom use, needle cleaning, HIV test procedures, and disease progression. Reliability was satisfactory; a correlation between test performance and condom use among sexually active teens provided predictive validity. (Contains 4 tables, 1 figure, and 4 notes.)