On the basis of anecdotal instances in which atypical laboratory findings suggested the possibility of unsuspected cryoglobulinemia, we applied predetermined criteria to determine how often such findings predict the presence of clinically significant cryoglobulinemia. The laboratory criteria are smeared M-spike (paraprotein) in agarose gel serum protein electrophoresis, precipitated protein at the serum application point of agarose electrophoresis gel, greater than 50% quantitative discrepancy between the densitometrically estimated M-spike and the relevant corresponding serum immunoglobulin isotype concentration from the same specimen, and smeared protein observed on an agarose electrophoresis immunofixation gel. Cases that fulfilled any of these criteria were prospectively collected for 2 years. Brouet types and clinical findings were determined in cases where cryoglobulins were subsequently identified and clinical data available. Among 83 patients in whom any of the above laboratory findings were identified, 52 had subsequent cryoglobulin evaluations. Fourteen of these 52 patients had cryoglobulinemia. Findings indicative of clinically significant cryoglobulinemia were present in 8 of the 10 patients in whom follow-up clinical data were available.