Lab experiments are all well and good, but they do have their limitations. It is necessary to study plants outside, in the field – or rather below the surface! This theme was explored at the UKPSF conference by Nick Ostle (Centre for Ecology and Hydrology, Lancaster) who argued that plant–soil ecological interactions are a critical – but poorly understood! – determinant of how land ecosystems respond to global changes. His talk extended the concerns beyond the purely biological to the wider issues surrounding ecosystem services, and even introduced the intriguing notion of ‘grassland biotechnology’(!). Linked to this was Sasha Mooney (University of Nottingham) looking (literally!) at root growth, but root growth with a difference: actually examining the growth of roots in the soil. The technique used was X-ray microcomputed tomography (μCT), which allows non-invasive imaging of intact roots in situ. Whilst using CT to image plant material is not new, the breakthrough is the computer program used to assemble and interpret the images recorded – RooTrak, which permits automated recovery of 3-D plant root architecture. Too many experiments on root growth and development are performed on roots growing in agar or on a lab bench, and about as far away from soil as you can get (and whose relevance to the real world is therefore questionable). So, this is an unparalleled opportunity to get the root’s perspective on life underground. As organs that support the above-ground part of the plant, that often store much of the photosynthetically generated biomass, that allow exploitation of nutrients and water reserves in the soil, and that form intimate and mutually beneficial associations with fungi and bacteria, etc (!!), the more we know about these secretive subterranean ‘systems’ the better. Accordingly, ‘RooTrak supports the computation of a range of quantitative measures and promises to facilitate future root phenotyping for trait-based crop breeding efforts’. I’m looking forward to the day when it will generate high quality images of a root and its associated mycorrhizal mycelial network – now, there’s a challenge!