To study patient's follow-up after finishing participation in randomized clinical trials (RCTs), and to identify factors associated with loss to follow-up (FU).Medical charts of 212 rheumatoid arthritis (RA) and osteoarthritis (OA) patients from a rheumatological out-patient center were analyzed. Loss to FU was considered when patients did not return to their regular appointments within the first year after finishing their participation in an RCT assessing anti-cyclooxygenase-2 non-steroidal anti-inflammatory drugs (anti-COX-2 NSAIDs). Mann–Whitney U-test, χ 2 test and Wilcoxon test were performed as appropriate. Logistic regression was performed to identify factors which might be related to loss to FU. A survey was conducted to obtain lost to FU patients' opinions. p values less than 0.05 were considered significant.The mean frequency of patients' visits in the year before enrollment in an RCT was 3.73 SD 2.06, and during the year after participation was 2.6 SD 1.96 (p<0.0001). Fifty patients (23.6%) did not return to their usual rheumatologic visit. On multivariate analysis, the number of daily tablets of study medication (odds ratio (OR)=2.64, 95% confidence interval (CI) 1.1 to 6.3) and the frequency of clinical visits (OR=0.56, 95% CI 0.37 to 0.85) were associated with loss to FU (p<0.008). Lost to FU patients' opinions did not support these findings.After participating in a RCT assessing anti-COX-2 NSAIDs, many patients return with less frequency, or do not return at all to their regular rheumatologic visit. Although a high number of tablets of the investigational drug and a low frequency of protocol visits may be contributors to patient loss to FU, investigators should consider that personal situations not related to the RCTs may also influence patients' return to consultation in the private setting.