Immunoglobulin replacement is a standard therapy for patients with primary immunodeficiencies. Subcutaneous administration of immunoglobulin offers more constant IgG levels than intravenous administration and simplifies administration for some patients. Use of L-proline as an excipient contributes to the stability of highly concentrated IgG preparations. The aims of the present study were to evaluate the pharmacokinetics of IgPro20 (Hizentra®), a new 20% subcutaneous IgG solution, and compare the area under the serum concentration-time curve (AUC) with that of a similar intravenous 10% IgG solution (IgPro10; Privigen®). At the request of the US FDA, an algorithm for determining IgG trough level ratios (TLRs) was developed in order to provide physicians with a practical tool for monitoring doses during steady-state IgPro20 therapy. This was a prospective, open-label, multicentre, single-arm, phase III clinical trial conducted in the US. The study was performed in a primary-care setting. Eligible patients were males or females aged 6-75 years with a primary immunodeficiency (common variable immunodeficiency or X-linked agammaglobulinaemia) who had received regular treatment with IgPro10 for at least 3 months prior to entering this study and had achieved serum trough concentration (C(trough)) values ≥5 g/L. IgPro20 was administered subcutaneously once weekly at initial doses equivalent to 130% of patients' previous doses, based on the results obtained in a Vivaglobin® study and due to an FDA request. After run-in, each patient's dose was adjusted to achieve an AUC comparable to that achieved with IgPro10 administered intravenously. Eighteen patients completed the study. Mean IgPro20 : IgPro10 dose ratio (dose adjustment coefficient) was 1.53 (range 1.26-1.87). The resulting mean AUCs were 105.6 g · day/L for IgPro20 versus 103.2 g · day/L for IgPro10 (geometric mean ratio 1.002; lower one-sided 95% confidence limit [CL] 0.951). Thus, the primary endpoint of the study was met, as this result exceeded the pre-specified criterion of the lower one-sided 95% CL of ≥0.8 for non-inferiority. At these AUCs, which were considered equivalent, the mean IgPro20 : IgPro10 TLR, determined by the developed algorithm, was 1.29 (range 1.18-1.73). Titres of specific antibodies tested were well above respective product specifications, suggesting that protection against infection would be effective. Steady-state AUCs with subcutaneous IgPro20 and intravenous IgPro10 were equivalent. Mean dose adjustment coefficient and mean TLR can be used for initial dose conversion without risk of under-protection but vary too widely to be considered measures of equivalence. Trial registration number (clinicaltrials.gov): NCT00419341.