With the growing use of immune checkpoint inhibitors (ICIs), case reports of rare yet life‐threatening pituitary‐adrenal dysfunctions, particularly for hypopituitarism, are increasingly being published. In this analysis, we focus on these events by including the most recent publications and reports from early phase I/II and phase III clinical trials and comparing the incidence and risks across different ICI regimens. PubMed, Embase, and the Cochrane Library were systematically searched from inception to April 2019 for clinical trials that reported on pituitary‐adrenal dysfunction. The rates of events, odds ratios (ORs), and 95% confidence intervals (CIs) were obtained using random effects meta‐analysis. The analyses included data from 160 trials involving 40 432 participants. The rate was 2.43% (95% CI, 1.73%‐3.22%) for all‐grade adrenal insufficiency and 3.25% (95% CI, 2.15%‐4.51%) for hypophysitis. Compared with the placebo or other therapeutic regimens, ICI agents were associated with a higher incidence of serious‐grade adrenal insufficiency (OR 3.19, 95% CI, 1.84 to 5.54) and hypophysitis (OR 4.77, 95% CI, 2.60 to 8.78). Among 71 serious‐grade hypopituitarism instances in 12 336 patients, there was a significant association between ICIs and hypopituitarism (OR 3.62, 95% CI, 1.86 to 7.03). Substantial heterogeneity was noted across the studies for the rates of these events, which in part was attributable to the different types of ICIs and varied phases of the clinical trials. Although the rates of these events were low, the risk was increased following ICI‐based treatment, particularly for CTLA‐4 inhibitors, which were associated with a higher incidence of pituitary‐adrenal dysfunction than PD‐1/PD‐L1 inhibitors.