Partial directed coherence (PDC) is a new frequency-domain method for determining the direction of information flow between multiple neural signals. It is based on the decomposition of multivariate partial coherences derived from multivariate autoregressive models. This approach is used to analyze the relationships between neural structures from simultaneous electrophysiological measurements. PDC reflects the concept of Granger causality, which describes the directional influence between variables. The paper discusses the generalization of directed coherence (DC) to multiple neural structures and introduces PDC as a method for structural analysis in the frequency domain. DC was originally developed to describe the directed coherence between pairs of structures, distinguishing between feedforward and feedback interactions. However, it was limited by the assumption that the covariance matrix Σ is diagonal. PDC overcomes this limitation by allowing for non-diagonal Σ, providing more accurate information about the direction of information flow. The paper also presents examples where PDC was used to reveal a reversal in the direction of information flow between the cortex and the hippocampus during a spindle episode in slow-wave sleep. PDC is a more general and accurate method for determining neural structure relations compared to traditional coherence analysis.Partial directed coherence (PDC) is a new frequency-domain method for determining the direction of information flow between multiple neural signals. It is based on the decomposition of multivariate partial coherences derived from multivariate autoregressive models. This approach is used to analyze the relationships between neural structures from simultaneous electrophysiological measurements. PDC reflects the concept of Granger causality, which describes the directional influence between variables. The paper discusses the generalization of directed coherence (DC) to multiple neural structures and introduces PDC as a method for structural analysis in the frequency domain. DC was originally developed to describe the directed coherence between pairs of structures, distinguishing between feedforward and feedback interactions. However, it was limited by the assumption that the covariance matrix Σ is diagonal. PDC overcomes this limitation by allowing for non-diagonal Σ, providing more accurate information about the direction of information flow. The paper also presents examples where PDC was used to reveal a reversal in the direction of information flow between the cortex and the hippocampus during a spindle episode in slow-wave sleep. PDC is a more general and accurate method for determining neural structure relations compared to traditional coherence analysis.