Introduction: Histo-anatomical structure of the heart is a key determinant of electrophysiological and mechanical cardiac function, and it affects their mutual interactions. Quantitative study of these interactions requires accurate knowledge of cardiac 3D structure across multiple scales, from sub-cellular levels to whole organ. Histological staining of sectioned tissue is the method of choice to provide high-resolution identification of cells and sub-cellular structures. However, 2D sections cannot easily be projected to 3D volumes, as they do not represent an inherently co-registered stack and show significant sectioning-induced deformations. Therefore, re-integration in 3D needs to be guided by additional data, such as magnetic resonance imaging (MRI) [1,2]. However, processing prior to cutting sections involves dehydration and embedding, which already cause significant changes in volume and shape of the wax-embedded sample, compared to the MRI images. This makes combining the data obtained with both methods challenging, so we introduced dual block-face imaging as an intermediate step in the pipeline for tissue reconstruction. Method: Briefly, rat hearts were excised after Schedule 1 killing, according to the UK Home Office guidance on the Operation of Animals (Scientific Procedures) Act 1986, and swiftly perfused using Tyrode solution (in [mM]: NaCl 140; KCl 5.4; MgCl2 1; HEPES 5; Glucose 10; CaCl2 1.8; pH 7.4). Hearts were then arrested using high-potassium Tyrode, and fixed with fast-acting Karnovsky’s fixative. Hearts were embedded in wax, mounted on a Leica RM2255 microtome and sectioned at 10μm. Two images of the wax block surface were taken prior to each cut with Matrix Vision USB 3.0 cameras (mvBlueFOX3): image 1 – surface-perpendicular projection, image 2 – Brewster angle projection via a linear polarization filter to obtain an image dominated by light reflected from surface areas that consisted of wax, not tissue. While the latter identified the precise pre-sectioning location of tissue contained in the next histological cut, the former corrects angular distortion to obtain true shape and location of the tissue. The mutual information guided subsequent 3D integration of digitally-imaged stained 2D stacks. Discussion: 3D histological reconstruction of extended tissue volumes at sub-cellular resolution with cell-type identification requires combined imaging methods to enrich in vivo reference (MRI), via identification of pre-cutting sample shape (dual block-face imaging), and cell type identification (stained 2D histology sections). Our improved system using a motorised microtome allows higher throughput efficiency of the collection of dual block-face data sets, required for effective 3D reconstruction of 2D histology image stacks.