Shoulder pain represents one of the most frequent clinical conditions in the general population and it can be generated by a large spectrum of pathologies. The most frequent approach of most rheumatologists to shoulder pain, in daily clinical practice, mainly based on their personal experience, is to locally inject drugs. Since the literature on this topic provides conflicting results due to the wide heterogeneity in the study designs, we decided to report the most relevant studies. Not enough data are available to assess whether US-guided injections are more efficient in controlling shoulder pain with respect to the landmark approach. However, it is likely that US-guided technique shows a more rapid improvement, possibly by providing a higher corticosteroid volume injected right where it is needed. When injecting hyaluronic acid, a more accurate localisation of the medicament might be useful to improve efficacy and avoid adverse effects (i.e. pain), however, there are no studies comparing the blind approach to the US-guided one. Finally, new treatments for shoulder pain have been used but they still need future validation in more appropriate RCTs.