Experimental models have suggested potential new treatments for human inflammatory neuropathy, but current practice is largely based on empirical trials. Evidence from randomized trials supports the use of intravenous immunoglobulin in Guillain-Barré syndrome, chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) and multifocal motor neuropathy with conduction block (MMNCB). In Guillain-Barré syndrome and CIDP intravenous immunoglobulin is equivalent to but more convenient than plasma exchange. In MMNCB adequate comparative studies of intravenous immunoglobulin and plasma exchange have not been performed. Corticosteroid treatment is beneficial in CIDP, but not in Guillain-Barré syndrome and may worsen MMNCB. More randomized trials and systematic reviews are needed to improve the evidence base for clinical practice.