Pressurized liquid extractions were performed on eight sediments in order to investigate if a modified US EPA method (100 °C, 100 bar, n-heptane/acetone (1:1), 2 × 5 min) provided exhaustive extractions of polychlorinated biphenyls (PCBs) from sediment, and to study if the extractability of PCBs from the different sediments was affected by characteristics of the sediment. The recovery from the eight native sediments, contaminated in nature, was between 96.4% and 98.9%, as an average of the recoveries from 10 PCB congeners. Hundred percent recovery was defined as the sum of two consecutive extractions (2 × 5 min each) at the stated conditions. The recoveries of the individual congeners were above 94%, except for one congener in one sediment, which had a recovery of 92%. When the recoveries and different characteristics of the sediments were compared, no correlation appeared between recoveries and sediment PCB concentration, total organic carbon (TOC), soot carbon (SC) or amorphous carbon (AC). The fact that carbon did not influence the extractions was somewhat surprising, since previous experiments have indicated a connection. Instead, statistically significant ( p < 0.05) correlations were observed for water content and carbon/nitrogen (C/N) ratio. The decrease in recoveries with decreased water content was attributed to less access of the solvent to the analytes due to less matrix swelling. The lowered recoveries with increased C/N ratio can indicate that a difference in structure of the organic matter exists, which influences the binding strength between the analytes and the matrix. The difference in structure can possibly be explained by different origin of the organic matter or by aging effects. Overall the method was found to be exhaustive and the excellent recoveries show that sediment characteristics do not influence the extractions markedly.