50 patients with subacromial pain were randomized to subacromial injection through a midlateral or posterior approach using anatomical landmarks. The primary outcome was rate of intrabursal injection, measured on radiograph using contrast medium administered with injection. Secondary outcomes included pain and shoulder function at 30 minutes after injection. Results demonstrated a significantly higher rate of accurate injection within the subacromial bursa in the midlateral group compared to the posterior group, though no significant differences in pain or function scores at 30 minutes after injection.