性侵害屬於極爲惡質之犯罪，且性侵害大都屬於嫌疑犯和被害人同處私密空間，及被害人未能及時報案採證，而增加偵辦及蒐證困難。本文針對國內地檢署及法院審理性侵害案採用測謊情形加以探討，以1999至2009年南部某警察局所實施「測謊鑑定」案件共660件，其中性侵害案計有142件；經法院判決計有467件，研究結果發現：一、實際測謊案件正確率即法院採信比率爲84.2%，測謊結果「誠實」被法院採信之比例爲94.9%，證實國內測謊鑑定之穩定及可靠。二、在「性侵害案」測謊結果被法院採信比例爲73.1%，與「非性侵害案」被採信比例爲86.9%比較，經卡方檢定呈顯著差異，顯示測謊於性侵害案中證明力可能被低估現象。三、測謊結果爲「誠實」被法院採信之比率94.9%，經卡方檢定顯著高於「說謊」被採信之比率74.8%，與先前研究相符，再次證實「測謊是確立無辜者之清白爲最大功用」。本文提出相關建議，提供法院增加測謊於性侵害案件證據之採用，以提升測謊證明力之參考。

In order to understand the affairs of polygraph examination application in Taiwan, the accurate rate of 660 real crime cases (142 sexual offense cases) were evaluated in South Taiwan from 1999 to 2009. The validity was assessed in the basis of polygraph examination outcome between the sexual and non-sexual offense cases. The results show that 84.2% of lie detection conclusions were accepted for the evidence in court. And the validity was 73.1% in sexual offense cases and 86.9% in non-sexual offense. The chi-square test was used to determine the validity differences between the sexual and non-sexual cases, and statistically significant were observed (χ^2=10.606, df=1, P＝0.001, P<0.05). Obviously, the probative proportion of sexual offense cases with polygraph examination was underestimated in Taiwan. The deception indicated group and no deception indicated group got admissibility rate in the courts of 74.8% and 94.9% respectively. This reveals that the largest function of polygraph is on innocent confirmation.