In this study, traditional Soxhlet, automatic Soxhlet and ultrasonic extraction techniques were employed to determine the speciation and concentration of polycyclic aromatic hydrocarbons (PAHs) on lime spray dryer (LSD) ash samples collected from the baghouse of a spreader stoker boiler. To test the efficiencies of different extraction methods, LSD ash samples were doped with a mixture of 16 US EPA specified PAHs to measure the matrix spike recoveries. The results showed that the spike recoveries of PAHs were different using these three extraction methods with dichloromethane (DCM) as the solvent. Traditional Soxhlet extraction achieved slightly higher recoveries than automatic Soxhlet and ultrasonic extraction. Different solvents including toluene, DCM:acetone (1:1 V/V) and hexane:acetone (1:1 V/V) were further examined to optimize the recovery using ultrasonic extraction. Toluene achieved the highest spike recoveries of PAHs at a spike level of 10 microg kg(-1). When the spike level was increased to 50 microg kg(-1), the spike recoveries of PAHs also correspondingly increased. Although the type and concentration of PAHs detected on LSD ash samples by different extraction methods varied, the concentration of each detected PAH was consistently low, at microg kg(-1) levels.