The presence of circulating tumor idiotype interferes with the in vivo effectiveness of anti-idiotype antibodies. We developed two assays that permit identification of patients with high levels of serum idiotype without the need for first producing an anti-idiotype antibody. A cell suspension made from the tumor was cultured for seven days with or without phytohemagglutin (PHA) and/or phorbol myristic acetate (PMA). Ig secretion in vitro by patients' tumor cells varied. In 4 patients, no secretion in vitro occurred, 5 patients had low levels, and 5 patients had high levels of Ig secretion. In three patients, Ig secretion occurred only after stimulation with PHA, PMA, or both. Spontaneous or induced immunoglobulin secretion in vitro is related to the levels of tumor idiotype secretion that exist in vivo. Eight patients with serum idiotype levels greater than 100 micrograms/mL (mean 265 micrograms/mL), had a minimum of 1.0 microgram/10(6) cells of idiotype secretion in vitro. Nine patients with serum idiotype levels less than 30 micrograms/mL (mean 3.7 micrograms/mL), had less than or equal to 0.5 microgram/10(6) cells of idiotype secretion in vitro. In another assay, the levels of IgM kappa and IgM lambda in patients' sera were compared with those in normal serum. An imbalance in the relative amounts of IgM kappa and IgM lambda indicated high levels of circulating idiotype in the serum, but this assay was less sensitive than the in vitro secretion assay and limited to IgM-secreting tumors. These assays will be useful for future clinical studies using anti-idiotype antibodies.