T he ability to clone primary tumors in soft agar has proven useful in the study of the kinetics and biological properties of tumor stem cells. We report the development of an in vitro assay which permits formation of colonies of human monoclonal plasma cells in soft agar. Colony growth has been observed from bone marrow aspirates from 75% of the 70 patients with multiple myeloma or related monoclonal disorders studied. Growth was induced with either 0.02 ml of human type O erythrocytes or 0.25 ml of medium conditioned by the adherent spleen cells of mineral oil-primed BALB/c mice. 5-500 colonies appeared after 2-3 wk in culture yielding a plating efficiency of 0.001-0.1%. The number of myeloma colonies was proportional to the number of cells plated between concentrations of 10⁵-10⁶ and back-extrapolated through zero, suggesting that colonies were clones derived from single myeloma stem cells. Morphological, histochemical, and functional criteria showed the colonies to consist of immature plasmablasts and mature plasma cells. 60-80% of cells picked from colonies contained intracytoplasmic monoclonal immunoglobulin. Colony growth was most easily achieved from the bone marrow cells of untreated patients or those in relapse. Only 50% of bone marrow samples from patients in remission were successfully cultured. Tritiated thymidine suicide studies provided evidence that for most myeloma patients, a very high proportion of myeloma colony-forming cells was actively in transit through the cell cycle. Velocity sedimentation at 1 g showed myeloma stem cells sedimented in a broad band with a peak at 13 mm/h. Antibody to granulocyte colony-stimulating factor did not reduce the number or size of the colonies. Increased numbers of myeloma colonies were seen when the marrow was depleted of colony-stimulating factor elaborating adherent cells before plating. This bioassay should prove useful in studying the in vitro biological behavior of certain bone marrow-derived (B)-cell neoplasia. In addition, systematic and predictive studies of anticancer drug effects on myeloma stem cells should now be feasible.