We present a robust, model-independent technique for quantifying changes in the dynamics underlying nonlinear time-serial data. After constructing discrete density distributions of phase-space points on the attractor for time-windowed data sets, we measure the dissimilarity between density distributions via L(1)-distance and chi(2) statistics. The discriminating power of the new measures is first tested on data generated by the Bondarenko "synthetic brain" model. We also compare traditional nonlinear measures and the new dissimilarity measures to detect dynamical change in scalp EEG data. The results demonstrate a clear superiority of the new measures in comparison to traditional nonlinear measures as robust and timely discriminators of changing dynamics. (c) 2000 American Institute of Physics.