Magnetic resonance imaging (MRI) has revolutionized the everyday management of spondyloarthritis, particularly in its axial forms. The ability of MRI to identify characteristic inflammatory changes not only at the sacroiliac joints, but also at the spine has shed new light on the diagnosis of spondyloarthritis and provided new pathophysiological insights, particularly regarding ossification of the spinal entheses [1]. Thus, sacroiliitis detectable only by MRI was included into the ASAS criteria set in 2009 [2], so that the radiographic sacroiliitis required in the modified New York criteria is no longer mandatory. This change may shorten the time to diagnosis [3] and constitutes official recognition of non-radiographic forms of axial spondyloarthritis [4], with a new terminology [5]. The unique capabilities of MRI have thus illuminated some of the obscure issues surrounding spondyloarthritis. Their discovery generated hope that many of the challenges raised by the early diagnosis or objective monitoring of spondyloarthritis (and therefore by the evaluation of treatments) might find “visible” solutions in everyday practice. Now, several years later, experience has tempered the initial enthusiasm. We must recognize that the studies conducted to date have raised new questions without necessarily solving the initial problems.