Background: Quality indicators (QI) are tools to evaluate the quality and effectiveness of laboratory testing. According to ISO15189:2012, QI’s are necessary for monitoring the performance through the different examination phases (pre-, post- and actual examination). The latter consists of the testing of patient samples, hence, is an important part of the medical diagnosis, treatment and patient monitoring process. From this perspective, we propose the ‘patient percentile monitoring’ project as QI for the examination phase. Methods: The project consists of the daily monitoring of medians of twenty commonly measured analytes in outpatients. All type and sizes of laboratories can participate. They are expected to calculate and send their medians to us in an automated way. We collect all information in our database, but after exclusion of weekend days. We then monitor the data by plotting of the moving median, but also the laboratories themselves can do it with help of a user-interface. We proposed preliminary limits for the assessment of the stability of performance. They are oriented on the biological variation, but, at the same time, respect the analytical reality. The laboratories are grouped by peer to allow instrument-specific comparison. Results: At the moment we have 64 participating laboratories with 115 different devices. We observe mid- to long term differences between different instruments, but also within-laboratory differences, sometimes accompanied by shifts or drifts. Another observation is that the moving median for certain analytes, e.g., C-reactive protein and gamma-GT, has higher variation in hospital laboratories. Focusing on outpatients seems promising for the assessment of laboratory bias. Conclusion: The patient percentile monitoring tool gives laboratories a direct, real-time QI to monitor the examination phase in compliance with ISO 15189:2012 accreditation requirements. We believe it will allow laboratories to better evaluate the mid- to long term stability of performance. Observed stability issues may be an incentive for root cause analysis, so that finally the tool may contribute to improved performance.