The effect of extraction variables, such as extraction time, temperature and extraction agent composition (X), in supercritical fluid extraction (SFE) of tributyltin (TBT) from a spiked sediment, has been optimized by using a factorial-fractional experimental design. Under the optimum conditions (T = 60 degrees C, P = 35 MPa, X = 5.1 M methanol in CO2, t = 30 min) the TBT extraction efficiency was 82% with a coefficient of variation (C.V.) of 9.2% (n = 5). Furthermore, the analytical procedure was validated using a reference material (RM) certified for its TBT content and was successfully applied to the analysis of a Certified Reference Material (CRM). Furthermore, an intercomparison of the SFE with another extraction procedure based on atmospheric pressure liquid extraction (APE) for the TBT determination in sediments has been carried out. Although TBT recoveries were significantly higher in the SFE procedure, the precision of both methods was comparable, probably associated to the similar steps in the extract treatment procedure (derivatization and clean up).