Measurement of S-100B protein in serum or cerebrospinal fluid is increasingly used as a molecular marker of brain damage. However, there is no information about the pre-analytical in vitro stability of S-100B. We have investigated whether storing blood samples at different time periods and conditions affected the measured levels of S-100B protein. Blood samples were taken from 29 neurosurgical patients with normal serum S-100B values and from seven patients with increased serum S-100B values. Blood samples were immediately divided into nine aliquots for measuring S-100B immediately and after 4, 8, 12, and 24 hours, stored at room temperature or at 4 degrees C. Measurement of S-100B was performed using the LIAISON assay (Byk-Sangtec Diagnostica, Dietzenbach, Germany). Moreover, in 10 additional patients the effect of freezing the serum and thawing the sample after 24 hours was investigated. There were no differences between the results of S-100B measurements after storing the sample at different temperatures and time periods. There was no trend towards higher or lower values in all three groups. Therefore, blood samples may be collected as part of the daily clinical routine without time constraint and even stored overnight without affecting S-100B serum levels when measured with the LIAISON Sangtec 100 assay.