Isrāʾīliyyāt narratives in classical tafsīr literature raise an unresolved tension between naqd al-sanad and naqd al-matan. When a report has a sound sanad yet contradicts Islamic doctrine, scholarly responses diverge. Some scholars preserve such reports through taʾwīl, while others reject them entirely. Muhammad Abū Shuhbah occupies the most uncompromising position in this spectrum, yet the epistemological basis of his rejection lacks systematic examination. Using a qualitative library research method, this study explains the epistemological foundation of Abū Shuhbah's rejection of Isrāʾīliyyāt reports with sound transmission chains and tests his argument's consistency through the Hārūt and Mārūt case. The primary source is Abū Shuhbah's Al-Isrāʾīliyyāt wa al-Mawḍū'āt fī Kutub al-Tafsīr, analyzed alongside relevant classical works via descriptive-analytical content analysis. Abū Shuhbah adopts the standard tripartite classification but treats a report as an indivisible whole. He rejects the Hārūt and Mārūt narrative due to an invalid matan, even when parts of its sanad are considered sound, and provides an alternative interpretation of Q. al-Baqara [2]:102 containing no Isrāʾīliyyāt elements. The soundness of a sanad cannot rehabilitate a report whose matan contradicts textual and rational evidence. This principle contributes to qawāʿid al-tafsīr by clarifying the structural relationship between naqd al-sanad and naqd al-matan, reframing Abū Shuhbah's stance as a principled epistemological position that assigns naqd al-matan the role of final arbiter.