Conducting randomised clinical trials (RCTs) in idiosyncratic drug-induced liver injury (DILI) is challenging. This systematic review aims to summarise the design and findings of RCTs in the prevention and management of idiosyncratic DILI. A systematic literature search up to January 31, 2020 was performed. Recognised scales were used to assess methodological bias and quality of the studies. Quantitative and qualitative analyses were performed. Heterogeneity was assessed with I statistic. Overall, 22 RCTs were included: 12 on prevention (n = 2,471 patients) and 10 in management (n = 797) of DILI/non-acetaminophen DILI-related acute liver failure (ALF). Silymarin (eight studies), bicyclol (four), magnesium isoglycyrrhizinate (three), N-acetylcysteine (three), tiopronin (one), L-carnitine (one), and traditional Chinese medicines (two) were tested in the intervention arm, while control arm mostly received standard supportive care or placebo. Main efficacy criteria in the prevention RCTs was DILI incidence or peak of liver enzymes value. In management RCTs, the efficacy parameter was usually 50 % decrease or normalisation of liver enzymes, or survival rate in DILI-related ALF patients. Overall, 15 trials described the randomisation method, eight were double-blind (n = 672) and nine had sample size estimation (n = 880). Four RCTs involving 377 patients used an intention-to-treat analysis. Based on the scarce number of trials available, tested agents showed limited efficacy in DILI prevention and management and a favourable safety profile. In conclusion, heterogeneity among studies in DILI case qualification and methodologic quality was evident, and the RCTs performed demonstrated limited efficacy of specific interventions. International research networks are needed to establish a framework on RCTs design and therapeutic endpoints.