New treatments against Alzheimer's disease (AD) may be just around the corner. A common approach in developing these disease-modifying treatments is to target beta-amyloid (Aβ). Aβ is excessively present in the AD brain and it likely starts to accumulate long before clinical symptoms become apparent. As Aβ is hypothesized to be the causative agent in the pathophysiological cascade leading to progressive neurodegeneration in AD, efforts to e.g. prevent its formation, to promote its clearance from brain tissue, and to inhibit its toxicity, are warranted. This quest for an effective AD treatment needs valid biomarker outcome measures, for instance because clinical benefit takes long to present itself and is difficult to measure, and also because treatment would likely be most efficacious if administered already before symptoms occur. In vivo amyloid imaging has evolved in the past decade to be a feasible means to monitor brain Aβ deposits in the human brain. It effectively differentiates AD patients from healthy age-matched controls, and also shows promise in the early, even presymptomatic, detection of AD. Amyloid imaging will likely also broaden and deepen our understanding of AD and other neurodegenerative disorders. It could prove valuable e.g. in subject selection and stratification for clinical trials, in safety and proof-of-concept assessments, and in monitoring of treatment effects. This article aims to review the motives, prerequisites, potential, and challenges of using amyloid imaging as a surrogate marker in clinical therapeutic trials in AD.