Low extraction efficiency (60-81%) of okadaic acid (OA) and dinophysistoxin 1 (DTX1) was obtained for 4 out of 5 shellfish species from Washington State (WA), USA, during application of a standard extraction method for determination of lipophilic marine biotoxins by LC-MS/MS as recommended by the European Union Reference Laboratory for Marine Biotoxins (EURLMB). OA and total OA including esters, DTX1, DTX2, and total DTX including esters, azaspiracid 1, 2, and 3 (AZA1, AZA2, and AZA3), pectenotoxin 2 (PTX2), and yessotoxin (YTX) were the toxins examined. Matrix-matched standards prepared from the same control samples used for spike-and-recovery tests were employed to evaluate toxin extraction efficiency and sample clean-up procedures. We adjusted the EURLMB extraction method by either using an acidified methanol extraction or pre-cooking shellfish homogenates at 70 °C for 20 min before EURLMB extraction. Extraction efficiency was improved markedly for OA and DTX1 with both modified methods and for YTX with the pre-cooking step included. However, recoveries were lower for YTX using the acidified methanol extraction and for PTX2 in non-mussel samples with the pre-cooking step. A hexane wash was applied to clean water-diluted non-hydrolyzed samples and a hexane wash was combined with solid-phase extraction for cleaning hydrolyzed samples. Improved sample clean-up, combined with LC-MS/MS adjustments, enabled quantification of U.S. Food and Drug Administration-regulated toxins in five shellfish species from WA with acceptable accuracy using non-matrix matched calibration standards.