Abstract: Information theory spread from original application in communication theory to many diverse fields, for instance physics (statistical mechanics), mathematics (probability theory) or computer science (algorithmic complexity). The basic ideas of information theory such as entropy, mutual information and entropy rate can be exploited in time series analysis to measure an information flow between two series, thereby detect possible causality. The information theoretical approach allows us detect statistical dependencies of all types, not only linear coupling as is the case for standard tools (auto/cross-correlation function, Granger causality test), hence, non-linear systems may be examined too.