背景與目的：下背痛須由理學檢查歸納出病理解剖問題，本篇即探討現今常用的理學檢查信度。
方法：收進2 位下背痛患者，兩位物理治療師分別進行理學檢查包括：觀察及觸診、動作檢查、
椎間活動度測試、及特別測試；受損模式分類包括：腰薦神經壓迫、軟組織受損、椎間盤突出、
小面關節問題、薦髂關節問題。卡帕檢定值（Kappa）大於0.40 表示信度中等以上。結果：觀察
及觸診判斷腰椎屈度信度中等以上（Kappa 值0. 3~0. ）；動作檢查之站姿前彎疼痛及躺姿重複前後彎疼痛具中等以上信度（Kappa 值0.4 ~0. ）；椎間活動度測試在第三至第五節的壓痛覺信度達中等以上（Kappa值0.42~0.9 ）；特別檢查除梨狀肌測試（Kappa值0.3 ）其餘都有很高信度（Kappa值0. 0~0.82）。而受損模式中除缺乏薦髂關節問題者，其餘皆具有中度以上信度（Kappa值0.43~0. 3）。結論：理學檢查項目中，特別檢查及引發疼痛症狀的項目會具有較高信度。即使不是各項檢查都具高信度，但受損模式仍可被一致歸類。臨床意義：臨床使用理學檢查時，需先定義清楚檢查方法及判定標準，方可增加檢查信度。（物理治療 2009;34（ ）:0000）

Background and Purpose: For designing proper treatment for the patients based on cluster of symptoms and sign related to anatomic-pathologic diagnosis, reliable physical examination (PE) is necessary. The object of this investigation is to examine the reliability of the common PEs for classifying LBP. Methods: Twenty-five subjects with LBP were recruited in this study. Two physical therapists conducted PE separately and the reliability of the PE and classification determined by the two examiners was analyzed afterwards. The PE included several parts: observation and palpation, movement test, spring test, and special tests; the final classifications based on PE and history of the patients included: nerve related symptom, soft tissue problem, disc related symptom, facet joint problem, and sacroiliac joint problem. The researcher conducted the reliability analysis and presented Kappa values which above 0.40 represent moderate level of reliability. Results: The reliability of observation and palpation shows substantial level in examining lordosis (0.53~0.71). In the movement test items, only standing flexion and repeated lying flexion reached moderate reliability (0.41~0.56). Provoking pain in spring test of L3~L5 reached moderate to excellent level of reliability (0.42~0.91). The special tests have moderate to excellent reliability (0.50~0.82) except for piriformis test (0.36). The classification also showed moderate to substantial level of reliability (0.43~0.63). No patients in our recruited population were classified as sacroiliac joint problem. Conclusions: The reliability of the items of special test and /or pain evoking tests was relatively higher than thers. The classification is identical even though not every item is reliable. Clinical Relevance: The examining items should be clearly defined with the examining procedure and standard before used clinically. (FJPT2009;34(1):29-41)