The properties and application of the median filter are investigated using both simulated and real data for signals evolving with time. Comparison is made with existing numerical techniques for drift compensation and noise reduction in analytical measurements such as the moving average and Savitzky–Golay digital filters. The median filter provides a means for dealing with "spiky" noise and separating peaks from a slowly changing baseline, even when the exact nature of the drift and noise distribution is not known. Median filtering is a useful and complementary addition to existing digital filtering techniques, being mathematically robust and readily implemented on any computer platform.Keywords: median filtering, signal noise, signal processing, digital filters, chemometrics, data smoothing.