本研究於加油站採樣取得178個土壤樣品，進行現場篩測儀器火焰離子化偵測器(FID)/光離子化偵測器(PID)篩測數據與實驗室氣相層析儀/火焰離子化偵測器(GC-FID)分析濃度之一致性檢核。現場FID篩測數據與實驗室(GC-FID)分析濃度相關性不佳，現場PID篩測數據亦呈現類似的結果。若現場篩測設定500 ppm為污染土壤的篩測基準，現場FID篩測判定有78個土壤樣品為污染土壤，現場PID篩測判定僅有28個土壤樣品為污染土壤。實驗室(GC-FID)分析TPH濃度以超過1,000 mg/kg為污染土壤的判定標準，則有56個土壤樣品歸類為污染土壤。FID篩測方式之偽陽性錯誤判斷約5.06%，PID篩測方式之偽陽性錯誤判斷約20.22%，此部分屬篩測合格的實際污染的土壤。 In this study, 178 soil samples were collected from gas stations to compare the consistency of on-site photoionization detector/ flame ionization detector (PID/FID) screening with lab analysis of soil total petroleum hydrocarbon (TPH) by gas chromatography coupled with flame ionization detector (GC-FID). The correlationship of on-site FID screening and GC-FID lab analysis is not good, and on-site PID screening presents the similar result. According to 500 ppm as on-site screening threshold, the classified contaminated soils were 78 soil samples by on-site FID screening and 28 soil samples by on-site PID screening. The TPH measurements by GC-FID lab analysis classified 56 soil samples as contaminated soils, in which TPH contents were more than 1,000 mg/kg. There exists 5.06% type-I error by on-site FID screening and 20.22% type-I error by on-site PID screening, which on-site screening is failed to correctly identify the real contaminated soils.