ObjectiveTo compare the pain expected to that effectively caused by magnetic resonance arthrography of the shoulder and, secondarily, to describe a simplified approach to the technique for articular access.Materials and MethodsWe prospectively evaluated 40 participants who used a visual analog scale and a simplified categorical scale to indicate the level of pain expected and that experienced after the procedure, comparing the two with the Wilcoxon matched-pairs test. We also determined gender-related differences in pain conditions using the Mann-Whitney U test. In addition, we described a modified technique involving radiographic localization and the use of standard puncture needles for articular access.ResultsAnalysis of the visual analog scales showed that the pain experienced was less than had been expected, with median scores of 1.75 and 3.75, respectively (p < 0.001). The level of pain expected was higher among women than among men, with median scores of 8.0 and 3.0, respectively (p = 0.014), as was the level of pain experienced, with median scores of 3.0 and 1.5, respectively (p = 0.139). The overall categorical evaluation corroborated that difference (p = 0.03). Articular access with the modified technique was successful in all patients.ConclusionMagnetic resonance arthrography of the shoulder is less painful than patients expect. In addition, digital radiographic guidance combined with the use of standard puncture needles appears to improve the efficiency of the method.