High-densities of ion channels in the axon initial segment (AIS) and nodes enable temporally precise action potential generation and conduction in vertebrate neurons (1). Computational models of large cortical pyramidal neurons show that the dimensions of the dendritic tree and the associated cable properties can impose a conductance and capacitive load on the initial segment causing axial lateral current loss into soma during the onset of the action potential. As a consequence of the reduced efficacy of sodium current to initiate action potentials it was recently hypothesized that the AIS ion channel density scales proportionally with the degree of dendritic conductance load (2). The passive cable properties of the AIS and its relationship to the dendritic tree are, however, not well understood. Here, we used a novel approach to obtain functional and structural information of single pyramidal neurons to examine whether features of the AIS scale with the dendritic tree. Wistar rats (n = 12) were anaesthetized (3% v/v isoflurane) and parasagittal brain slices (300 µm thick) prepared using standard methods. Whole-cell patch-clamp recordings were obtained from thick-tufted layer 5 pyramidal neurons filled with biocytin and quantified for action potential properties and the subthreshold transients (n = 23). Slices were subsequently fixed in 4% paraformaldehyde and stained for AIS-specific anchoring protein markers and biocytin-streptavidin allowing examination and identification of the entire axonal and dendritic tree in the slice. High-resolution confocal laser-scanning microscopy (Leica TCS SP8) was used to collect detailed information on the AIS and nodal domains together with structural information of dendrites and axons. The results indicated that the soma area positively correlated with the AIS area (r = 0.85, Pearson’s bivariate correlation coefficient, p < 0.001, n = 24). Structure-function correlation showed furthermore that the location, but no other aspects of the AIS correlated with the somatic action potential rate-of-rise (r = -0.72, Pearson’s p < 0.001, n = 24). To systematically explore the causal relationships we integrated immunofluorescence data into three-dimensional reconstructions allowing a detailed quantitative representation of the axonal domains into a morphologically realistic conductance-based computer model, constrained by the experimental data. Preliminary analyses indicate that both geometry-dependent and geometry-independent rules exist for the size and location of the axon initial segment. These data suggest that the geometry of the axon initial segment in part compensates for the capacitive and conductance load of the cell, leading to both normalization and tuning of the axo-somatic firing properties.