The anti-CD30 immunotoxin (IT) Ber-H2/saporin is effective in patients with refractory Hodgkin's disease. However, responses are short and partial, one of the main reasons being the inability to repeat IT doses because of formation of human antibodies against the murine antibody and/or the toxin. To overcome this problem, we constructed two new anti-CD30 ITs by covalently linking the mouse monoclonal antibody Ber-H2 to the type 1 ribosome-inactivating proteins (RIPs) momordin (MOM) and pokeweed antiviral protein from seeds (PAP-S), which do not cross-react with each other or with saporin. Both ITs inhibited protein synthesis by Hodgkin's disease and anaplastic large-cell lymphoma (ALCL)-derived CD30+ target cell lines with a very high efficiency (IC50 ranging from < 5 x 10(-13) M to 2.75 x 10(-11) M, as RIP). In a SCID mouse model of xenografted CD30+ human ALCL, a 3d treatment with non-toxin doses of Ber-H2/MOM (50%LD50), started 24 h after transplantation, prevented tumour development in about 40% of the animals and significantly delayed tumour growth rate in the others. Main toxicity signs in mice and rabbits were dose-related increase of serum transaminases (AST and ALT) and creatine phosphokinase (CPK). LD50 (as RIP) in Swiss mice was 7 mg/kg for Ber-H2/MOM and 0.45 mg/kg for Ber-H2/PAP-S. Sequential administration of two anti-CD30 ITs (Ber-H2/MOM and Ber-H2/saporin) was well tolerated and did not result in formation of antibodies cross-reacting and with the two plant toxins. The results presented in this paper suggest that in the future, sequential administration of anti-CD30 humanized antibodies linked to antigenically distinct type 1 RIPs (saporin, MOM, PAP-S) should be feasible.