"Laminar structure is a distinctive feature of cerebral cortex which comprises 6 layers, in which each layer is characterized by distinct cellular organization, specific inputs and projection targets. Anatomical circuitry across laminae has been well-studied, and previous studies have shown laminar-dependence of neural responses, suggesting that sensory information is processed in a laminar-dependent manner. However, the functional circuitry across laminae related to the underlying mechanism of laminar processing remains largely unexplored. This project is to explore functional connectivity measured with spiking activity across laminae, and test the hypothesis that functional connections triggered by first-order stimuli and second-order stimuli share great similarity. Here, by using a series of functional connectivity methods, we investigate association and causal relationship across laminae in primary visual cortex. While responding to visual stimuli, neurons across cortical depth in A17 of the anesthetized and paralyzed cats were recorded with a 32-channel linear array probes (NeuroNexus). Current source density (CSD) analysis was applied to low frequency components of sinewave grating responses, to approximately localize recording sites into 3 layers: (supra-granular (SG), granular (G), and infra-granular (IG)) (Nicholson & Freeman, 1975; Mitzdorf & Singer, 1978). Multiunit activity (MUA) was extracted from recorded neuronal responses for each channel, as times of level-crossings at 3 standard deviations of the bandpass digitally filtered (300 Hz - 3.0 kHz) signal. We implemented a series of functional connectivity methods in Matlab, including mutual information, Pearson correlation, Granger causality (GC) and conditional Granger causality to measure neural interactions between cortical layers with multiunit spiking activityMUA (MUA) data. Analysis of conditional Granger causality could remove variance explained by other variables when estimating pairwise functional connectivity. Here we used a nonparametric version of Granger causality analysis, designed to handle spiking responses (Y. Chen, Bressler, & Ding, 2006; Dhamala, Rangarajan, & Ding, 2008; Hirabayashi, Takeuchi, Tamura, & Miyashita, 2013; R. Chen, Wang, Liang, & Li, 2017a). For validation, we tested these methods on simulated spiking data from a small network of leaky integrate-and-fire models. Then we analyzed datasets triggered by both luminance-modulated (LM) static gratings and contrast-modulated (CM) static gratings and compared the patterns of functional connectivity. These methods showed patterns of functional connectivity across laminae, exhibiting a similar relation to the laminar structure measured by CSD. The causal influences between cortical layers, as measured by GC and conditional GC, are predominantly in the direction from the G layer to SG layers, from the SG and G layers to the IG layerAnalysis of GC and conditional GC indicated directionality of connections. The results of these two methods also suggests that inter-laminar connections were greatly strengthened by responses within intra-laminar connections. In conclusion, these three functional connectivity measures, including mutual information, Pearson correlation and Granger causality, suggest how neurons in the primary visual cortex process sensory inputs in a laminar-dependent manner. Functional connectivity triggered by first-order stimuli and second-order stimuli shares great similarity in pattern of directional biases, consistent with shared processing mechanisms for these two kinds of stimuli, though with some differences in strength of connections"--