Common variable immunodeficiency (CVID) is not a homogeneous disease, as has become clear from recent scientific studies. This makes the interpretation of studies of clinical therapeutics difficult to assess and raises questions about historical case reports. The evidence for the optimum use of replacement immunoglobulin in CVID is reviewed. This therapy represents the current gold standard, despite attempts to use other immunostimulatory compounds. Questions of product properties, product selection, adverse events and infectious risks are addressed. Products are not interchangeable and have different physicochemical characteristics. Despite intravenous immunoglobulin being in use for 20 years, there are still unanswered questions over dose and target trough IgG levels, particularly with respect to patients with established lung disease. The management of organ-based complications of CVID is discussed. This includes the treatment of unusual infections such as mycoplasmas and enteroviruses, which are specific to antibody deficiency. The diagnosis and treatment of the granulomatous disease of CVID is discussed. The role of surgery, including lung transplantation, in the management of CVID complications is reviewed. There are few available data on optimum strategies for antibiotic usage for bacterial infective complications and it is clear that present regimens, at least in severe recurrent sinus disease, are not consistently effective. Better clinical trials are required to identify appropriate regimens and validate or disprove widely held assumptions about therapy in CVID. Despite advances in diagnosis and management, there is abundant evidence in the UK that patients do not yet receive rapid diagnosis and optimum therapy, even within the limited published data currently available. This leads to considerable avoidable morbidity and mortality.