A beta 2-microglobulin (beta 2m) 'modifying activity' has been demonstrated by crossed radioimmunoelectrophoresis of serum. The activity could be estimated by planimetry and expressed in arbitrary units (A.U.). Elevated values of beta 2m 'modifying activity' (greater than 0.30 A.U.) has been demonstrated in 49 of 54 patients with small cell lung cancer. The values returned to normal (less than 0.30 A.U.) in eight of the ten patients achieving complete remission (CR) and in three of seven patients achieving partial remission (PR) after chemotherapy. The decrease was more pronounced (median 0.56 versus 0.16, p less than 0.01) in patients achieving CR compared with patients achieving PR. Relapse was accompanied by rising values in 11 of 15 patients monitored during chemotherapy. In nine of these patients abnormally high values of beta 2m 'modifying activity' was demonstrated more than 1 month before clinical or radiological evidence of disease progression. Total serum beta 2m was measured by radioimmunoassay. Elevated values (greater than 200 nmol/l) was found in 14 of 48 patients with small cell lung cancer. No correlation with the clinical course was found in patients monitored during chemotherapy. Estimation of total beta 2m is of no clinical value in small cell lung cancer. Estimation of beta 2m 'modifying activity, provides clinically relevant information, but is too laborious for routine clinical application. The biochemical process underlying this phenomenon should be studied further to allow development of a more simple test.