Purpose MRI scanner performance temporal stability is fundamental aspect in several multi-parametric exams (like diffusion, functional and spectroscopy) effect of order of few percent may be clinically relevant. Revised version of Friedman protocol (JMRI, 23 -2006) was adopted to check stability of scanners over time on weekly basis. Linewidth of water peak in MR spectroscopy after optimization is considered as well. Results over a long period survey of several clinical scanners are reported. Methods The purpose of the procedure is to measure time stability of the scanners in conditions that are as similar as possible to the clinical settings. The head coil and scanning sequence used clinically were tested. A time series of 200 images of QA vendor phantom is collected. The automated analysis provides information about signal intensity, uniformity, magnitude spectrum, radius of decorrelation (RdC), short term reproducibility, SNR and signal fluctuation to noise ratio. RdC represents the size of ROI for which the main source of signal noise is systematic and not statistic. With same setup, linewidth after B0 homogeneity optimisation is measured. Data were acquired on four 1.5 T MRI clinical scanners, on a 4 years time interval. Results Data fluctuations of considered parameters are of order of few percent (varying between 1% and 9%), significant variations have been observer after component change and software upgrade. Fourier analysis of short term signal doesn’t show periodic noise. Short term reproducibility is below 1% for all considered scanners. RdC is 18 pixels for two scanners and 6 pixels for other. No significant variation have been observed during temporal series. Linewidth is stable within 6%. Conclusions Regular check of parameter stability provides strong feedback regarding scanner performances, and may allow to predict possible malfunctioning. Proposed protocol may be used as basic QA in advanced MRI technique.