The field of quantitative microscopy (QM), as applied to base-metals, alloys and their products, is considered to be mature, sophisticated, and essential to the manufacturing industry. However, the role of QM in the mining sector is still developing. Compared to man-made metals and materials, natural structures such as ore bodies can be far more complex and varied, and difficult to characterize at the microstructural level. This has limited the scope of QM in petrography until recent times. Now, enormous possibilities have opened up due to the availability of powerful computers, automated microscopes, image analysis software and advanced process modelling techniques. In most mining applications, representative sampling and specimen preparation procedures for QM need to be far more rigorous than those familiar to metallographers. It is arguable that the success or failure at the industrial level of many new applications will depend on the development of practical methodologies to meet these high demands. This paper covers some of the challenges and pitfalls of automated petrography and some improvements in methodology developed with guidance from established standards in metallography.