MDAY-D2 is a highly tumorigenic and anaplastic DBA/2 strain murine transplantable tumor capable of rapid and widespread spontaneous metastatic growth. It was therefore chosen as an ideal murine tumor model for the study of factors affecting metastatic growth. Two approaches were taken in an effort to obtain stable qualitative and quantitative low-metastatic variants of MDAY-D2, namely, cloning of multiple sublines and derivation of lectin-resistant (LecR) mutants. In the first method, 20 clones were isolated, and of these, three initially showed a marked reducstion in ability to metastasize from a subcutaneous site. However, these clones proved to be unstable both in vivo and in vitro. In the LecR selection experiments, 18 independent variants were obtained using chemical mutagenesis followed by treatment with wheat germ agglutinin (WGA), phytohemagglutinin (PHA), or concanavalin A (Con A). All of the variants proved to be highly metastatic except two WGAR variants, designated MDWI and MDW3. They proved to be nontumorigenic in normal DBA/2 hosts even when as many as 5 X 10(6) cells were injected, and this was found to be a stable change. Despite this fact, the nontumorigens an unchanged expression of H-2d and Ly-6.2 alloantigens and Fc receptors. The variants were, however, tumorigenic and metastatic in severely immunosuppressed (nude) mice, but not in moderately immunosuppressed 250-R-irradiated DBA/2 hosts. The results demonstrate that 1) stable membrane mutant sublines possessing radically altered growth properties in vivo can occasionally be obtained by selection of LecR variants, and 2) their growth and metastatic properties can be greatly affected by the immunologic status of the host. The possibility that the chemical mutagen treatment itself induced, or was responsible for, MDW1 and MDW3 variant formation is also discussed.