The primary aim of this study was to evaluate the prevalence of opioid prescriptions in hospitalized geriatric patients. Other aims were to evaluate factors associated with opioid prescription, and whether or not there was consistency between the presence of pain and prescription. Opioid prescriptions were gathered from the REgistro POliterapie Societa` Italiana di Medicina Interna (REPOSI) data for the years 2008, 2010 and 2012. 1,380 in-patients, 65+ years old, were enrolled in the first registry run, 1,332 in the second and 1,340 in the third. The prevalence of opioid prescription was calculated at hospital admission and discharge. In the third run of the registry, the degree of pain was assessed by means of a numerical scale. The prevalence of patients prescribed with opioids at admission was 3.8% in the first run, 3.6% in the second and 4.1% in the third, whereas at discharge rates were slightly higher (5.8, 5.3, and 6.6%). The most frequently prescribed agents were mild opioids such as codeine and tramadol. The number of total prescribed drugs was positively associated with opioid prescription in the three runs; in the third, dementia and a better functional status were inversely associated with opioid prescription. Finally, as many as 58% of patients with significant pain at discharge were prescribed no analgesic at all. The conservative attitude of Italian physicians to prescribe opioids in elderly patients changed very little between hospital admission and discharge through a period of 5 years. Reasons for such a low opioid prescription should be sought in physicians' and patients' concerns and prejudices.