Published reports dealing with the effects of man-made environmental electromagnetic fields (EMFs) (powerlines and mobile phones, as examples) on human brain electrical activity have been inconsistent. We hypothesized that the problem arose from the widespread use of linear methods (time averaging, spectral analysis) to analyze what were essentially nonlinear stimulus-response relationships. We used a nonlinear analytical method to detect deterministic changes in brain electrical activity induced by weak magnetic fields (2 G, 60 Hz).