OBJECTIVES: Computer Adaptive Testing (CAT) are assessment instruments based on the Item Response Theory. CAT use a reduced number of questions (selected from a calibrate item-bank) in evaluating the subject with a minimum error. These are characteristics which would be very useful in the Health-Related Quality of Life-HRQoL area. The purpose of the present work was to develop a calibrated bank of HRQoL items to used it in the to construction of a CAT for evaluating the generic HRQoL in our country: the CAT-Health. METHODS: HRQoL was considered as an unidimensional construct. An expert panel constructed the item bank with items from validated HRQoL questionnaires and original ones, all with a likert scale of five alternatives. A pilot group was interviewed with all these items to evaluate parameters of Classic and Item Response Theories and select the anchoring items. A normative group answered part of the items (through a computer application or in paper) and items were re-calibrated (using Samejima's Graded Response Model-GRM) and norms of item selection were designed. Various simulation studies were carried-out with the final calibrated bank of HRQoL items. RESULTS: Experts panel constructed a bank of 140 HRQoL items. Pilot group (N=185) answered all the items: 12 items were eliminated because homogeneity coefficient was under 0.20; Cronbach's alpha was 0.99; factorial analysis confirmed the uni-dimensionality of the item bank; IRT showed that 23 items had a “b” parameter which did not fit the model (eliminated). Normative group (N=1.373) answered 10 different questionnaires (5 anchoring items and 10 different) constructed with the 105 items remaining in the bank: Information function indicated that 9 items must be eliminated; GRM was applied to re-calibrate the items. Finally, norms of item selection were designed and simulations were carried-out. CONCLUSION: A new CAT to evaluate generic HRQoL has been developed and tested through simulation techniques.