Cholinesterases were characterized in the serum of 77 treated and 11 untreated patients having primary carcinomas of various tissue origins and 21 healthy volunteers which served as controls. In most of the samples, pseudocholinesterase (BuChE) accounted for almost all cholinesterase (ChE) activity and was inhibited by the organophosphorous poison tetraisopropyl pyrophosphoramide (iso-OMPA). In samples from the tumor-bearing patients, ChE degraded 733 +/- 59 nmole acetylcholine/h/mg protein, lower than the 960 +/- 175 nmole/hour/mg levels measured in controls. Tumor serum ChE exhibited elevated sensitivity to 1,5-bis-(4-allyldimethyl ammonium phenyl)-pentan-3-one dibromide (BW), the selective bisquaternary inhibitor of "true" acetylcholinesterase (AChE), with no correlation to age, sex, staging of tumor, presence of metastases or the specific treatment protocol, and with a different distribution pattern from the decrease in ChE specific activity or the sensitivity to iso-OMPA. In sucrose gradients, ChE sedimented as 12S in controls whereas in tumor serum samples from treated patients an additional component of 6 to 7 S, inhibited by both iso-OMPA and BW, also was detected. However, the ChE activity in serum of patients with diagnosed carcinomas before surgery and medical treatment appeared to be nondistinguishable from controls. These findings suggest that the modified properties of serum cholinesterases in carcinoma patients are not the result of the tumor itself, but that the common therapy protocols used in the treatment of primary carcinomas may cause the appearance of soluble ChE activity with properties of both AChE and BuChE, which accumulates in the serum.