Ovarian cancer has a high mortality and novel-targeted treatment strategies have not resulted in breakthroughs for this disease. Insight into the molecular characteristics of ovarian tumors may improve diagnosis and selection of patients for treatment with targeted therapies. A potential way to achieve this is by means of molecular imaging. Generic tumor processes, such as glucose metabolism (F-18-fluorodeoxyglucose) and DNA synthesis (F-18-fluorodeoxythymidine), can be visualized non-invasively. More specific targets, such as hormone receptors, growth factor receptors, growth factors and targets of immunotherapy, can also be visualized. Molecular imaging can capture data on intra-patient tumor heterogeneity and is of potential value for individualized, target-guided treatment selection. Early changes in molecular characteristics during therapy may serve as early predictors of response. In this review, we describe the current knowledge on molecular imaging in the diagnosis and as an upfront or early predictive biomarker in patients with ovarian cancer.