In adults with alcohol‐related liver disease, glucocorticosteroid (GCS) and pentoxifylline, both given alone, were compared with placebo/no intervention; GCS plus pentoxifylline was compared with placebo, GCS alone, or pentoxifylline alone; and pentoxifylline was compared with GCS. Randomized controlled trials reported no clear differences in mortality or the incidence of decompensated cirrhosis for any of these comparisons. More people experienced adverse events with GCS alone or with pentoxifylline alone than with placebo/no intervention (on average, 572 vs 143 per 1000 people, and 373 vs 212 per 1000 people, respectively). Investigators did not report the preferred measure of the number of people experiencing adverse events for the other comparisons; the difference in the number of adverse events experienced by participants cannot be translated into an increase in the number of people experiencing adverse events and therefore is limited in its clinical application. When assessed, the quality of the evidence was very low, and no firm conclusions can be drawn.