Hybrid PET/MRI is the obvious next step towards integration of multiparametric imaging data into clinical medicine and medical research. It provides structural, functional and biochemical information at the same time. To date, a few years after the commercial launch of the first generation of scanners, the technique is in its infancy, and several technical hurdles still need to be dealt with. Efficient workflows eliminating redundant information are required to get the best out of the two worlds without compromising the quality of the individual components. Integration of PET and MRI science requires an additional effort but at the same time this process nicely fits in the current perspective of an integrated imaging profession. In this article, the authors provide an overview of the current clinical data on PET/MRI and a perspective on some potential applications in the near future, with a focus on oncology, neurology, cardiovascular diseases and rheumatology. Tijdschr Nucl Geneesk 2013; 35(4):1160-1171 Introduction Only a decade after the successful introduction of positron emission tomography combined with computed tomography (PET/CT), integration of PET with magnetic resonance imaging (PET/MRI) was introduced clinically as the next hybrid imaging option. Since these two modalities (i.e., PET and MRI) are not easily combined in a single scanner, its design has taken years of technical research. To date, there are two options: an integrated system housing both components in a single gantry, or a geometrically separated PET and MRI on either side of a rotating table. The former has the obvious advantage of truly simultaneous image acquisition, and the latter does not require any concession for either component. Clinically, the paradigm change towards individualised 'tailored' therapy leads to an increased need for routine diagnosis at a molecular level. Most molecular biology methods mandate tissue sampling for in vitro analysis. In contrast, molecular imaging allows for non-invasive studies at the molecular level in living, intact organisms. Medical imaging can provide anatomical, biological and pharmaceutical information in individual patients, accounting for heterogeneity at each of these levels, non-invasively and at a quantitative level. Therefore medical imaging is a logical link between in vitro sciences and clinical medicine. However, no current imaging modality can cover the full spectrum of data required for personalised medicine by itself: PET only measures a single molecular process at the same time, and MRI has significant limitations with respect to molecular contrast. With respect to PET/MRI systems versus PET alone systems, MRI data adds functional parameters to the PET information (1,2), including cellular density (using diffusion-weighted imaging (DWI)), angiogenetic and perfusion characteristics (using intravenous contrast; dynamic contrast-enhanced (DCE)), tractography in cerebral white matter, real-time cardiac contractility and movement, metabolic data by spectroscopy, information on oxygenation (blood-oxygen-level-dependent (BOLD) imaging), etc. Moreover, MRI has some obvious advantages versus CT, including lack of radiation exposure and superior soft-tissue contrast. Whilst in the case of PET/CT, the merged modality quickly found its way in routine daily practice, PET/MRI is facing a number of technical challenges that never were an issue for PET/CT (3). These issues require a high level of multidisciplinary collaboration in order to capitalise the full multiparametric potential of PET/MRI. However, the current challenge is to make PET/MRI suitable for clinical practice. To some extent, implementing PET/MRI is also challenging for its individual components: for PET to validate new radiotracers demonstrating differences in cancer phenotypes between and within patients, and for MRI to achieve proper platform-independent standardisation. Current clinical applications with PET/MRI are still in their infancy. Here we provide an overview of some of its potential in the clinical domains of oncology, neurology, cardiovascular diseases and rheumatology.