Objective This post hoc analysis evaluated romiplostim self-administration (SA group) vs. romiplostim administration by a healthcare provider in a clinical setting (HCP group) in patients with chronic immune thrombocytopenia (ITP). Methods Outcomes from 3 ITP trials allowing self-administration in patients achieving a stable romiplostim dose for ≥3 consecutive weeks were compared. Evaluations were conducted for 12-wk treatment intervals. Efficacy endpoints included percentage of patients and weeks with platelets within the target range of 50–200 í 109/L and safety. Results Baseline characteristics suggested less severe disease in the SA groups (n = 563) than in the HCP groups (n = 241). The SA groups had greater proportions of patients achieving the target platelet range (55–58% vs. 40–52%) and greater proportions of weeks with a platelet response (75–88% vs. 47–76%) than the HCP groups. The rate of romiplostim discontinuation was twofold to fivefold lower in the SA groups than in the HCP groups. Rates of duration-adjusted adverse events (AEs), serious AEs and treatment-related AEs were also lower in the SA groups. Conclusions In conclusion, in adults with ITP receiving romiplostim, self-administration was comparable to healthcare provider administration in terms of efficacy and safety profiles, suggesting that self-administration of romiplostim is a feasible option for certain patients with ITP.