Polypharmacy and the consequential risks of drug−drug interactions (DDIs) have been well documented [1] and have also been described in HIV infection [2,3]. As our patients become older with comorbidities, the number of coprescribed medications increases. So-called 'polypharmacy', usually defined as five or more medications, has been linked to decreased adherence, increased adverse drug reactions, more hospitalizations and an increase in mortality [4]. Evans et al. recently found that 27% of HIV-infected patients had a clinically significant drug interaction between their highly active antiretroviral therapy (HAART) and another agent [5].We performed an audit to quantify polypharmacy in our large UK HIV-infected cohort using an HIV clinical care database. Using our HIV clinic database for patients attending in the year prior to 4 October 2013, we extracted demographics, antiretroviral regimen and list of all other medications. We looked for evidence of potential DDIs, of which we present three here.Our cohort comprised 1415 patients, of whom 92 (6.5%) were aged over 60 years. Among 1395 patients with evaluable data, 1227 (88%) were on HAART and 988 (71%) reported taking other medications. The median number of coprescribed drugs was 2 [range 0 to 27; interquartile range (IQR) 4], with 21% of patients taking more than five drugs. A total of 574 different medicines were recorded (Table 1). The median number of antiretroviral formulations taken was 2 (range 0 − 5), with 303 (24% of treated) patients on a single tablet regimen. Thus, the total number of medications taken by our cohort was a median of 4 (range 0-28). In patients aged over 60 years, and in those with hepatitis C virus coinfection, the median number of coprescribed drugs rose to 4 [range 0-27 (IQR 1) and range 0-17 (IQR 2), respectively], and with the addition of their ARV regimen this rose to 6 total medications.In spite of this, we found little evidence of critical DDIs. No patients were found to be taking rilpivirine together with a proton pump inhibitor (PPI). One patient was recorded as taking a contra-indicated statin together with a protease inhibitor (PI); however, on case note review this interaction was found to have already been intercepted and a more appropriate statin prescribed. There were 78 patients identified who were coprescribed steroid preparations together with a PI. However, only seven of these were considered to be at risk of harm that had not been assessed and considered by the clinician.This audit highlights the ongoing potential for serious DDIs in HIV practice as a result of the degree of polypharmacy. Steroid coprescription is a concern, but two other concerning DDIs were not prevalent. Our findings rely on good data recording as well as a cooperation in good medicines reconciliation from both the clinician and the patient. We are introducing routine checking of our emergency care summary (ECS: an online record of prescriptions) in our out-patient clinical practice. In addition, we have asked primary care to add hospital-pres...