Aging of the population represents one of the main challenges for health systems because of the increase in the demand for hospital services. To be able to count on tools that allow an objective evaluation of hospital-resource use becomes indispensable for health systems. To evaluate the reliability and validity of the Appropriateness Evaluation Protocol (AEP) regarding the appropriateness of admissions and hospital stays in elderly patients. In a scenario of scarce resources, to have a valid instrument will make it possible to evaluate the process of care in our growing elderly population in a standardized way. We carried out a retrospective study of 144 randomly chosen elderly patients admitted to the hospital with 394 even-numbered hospital-stay days. For the reliability analysis between the pair of nurses with the AEP and the pair of specialists, the details of the hospital admissions and the stay days were obtained from the clinical files. Criteria validity was conducted by pairs of physicians, including two internists, two general surgeons and two geriatricians. Only the agreements were compared with agreements of the AEP-trained nurses. Disagreements were excluded from the final analysis. Inter-rater (inter-reviewer) agreement of hospital admissions and days spent by the patient presented a kappa coefficient of >0.70, while these admissions and hospital-stay days was >0.70. Sensitivity and positive predictor value to detect inappropriate admissions were not calculated because no agreement existed on inappropriate admissions. Specificity and negative predictive value to detect appropriate admission was >94.0% and >98.0%. Sensitivity and positive predictor value to detect inappropriate hospital-stay days was >44.0% and >10.0%, while specificity and negative predictor value for detecting appropriate hospital-stay days was >79.0% and >88.0%. AEP's high-reliability and moderate-validity results with regard to clinical judgement positions it as a useful instrument for appropriate hospitalization screening in elderly patients.