We measured perioperative plasma concentrations of brain-derived neurotrophic factor (BDNF), a major mediator of synaptic plasticity in the central nervous system, in males, 30-65 years old, undergoing lumbar or cervical discotomy. Patients were randomly allocated to a general anesthetic with propofol induction and maintenance or with thiopental induction and isoflurane maintenance. BDNF plasma concentrations were measured before induction (baseline), 15 min after induction but before start of surgery, at skin closure, in the post-anesthetic care unit, and 24 h postoperatively. Data from 26 patients (13 in each group) were analyzed. At each time point, BDNF plasma concentrations showed large variability. At baseline, concentrations were 631 +/- 337 (mean +/- SD) pg ml(-1) in the propofol group and were 549 +/- 512 pg ml(-1) in the thiopental-isoflurane group (P = 0.31). At 15 min, concentrations significantly decreased in the propofol group (247 +/- 219 pg ml(-1), P = 0.0012 compared with baseline) but remained unchanged in the thiopental-isoflurane group (597 +/- 471 pg ml(-1), P = 0.798 compared with baseline). At skin closure and in the post-anesthetic care unit, concentrations were not different from baseline in both groups. At 24 h, concentrations significantly decreased below baseline in both groups (propofol: 232 +/- 129 pg ml(-1), P = 0.0015; thiopental-isoflurane: 253 +/- 250 pg ml(-1), P = 0.016). In the propofol group, there was a weak but statistically significant positive correlation (R2 = 0.38, P = 0.026) between the duration of surgery and BDNF plasma concentrations at skin closure. These data suggest that in males undergoing elective minor surgery, BDNF plasma concentrations show a specific pattern that is influenced by the anesthetic technique and, possibly, by the duration of surgery.