Apatite is a common accessory mineral in U–Pb geochronology and provides valuable constraints on magmatic evolution, orogenic processes, and basin thermal histories. However, its generally low U concentrations and high common Pb contents have limited its broader application. The long-term stability of matrix-matched reference materials and the reliability of common Pb correction are critical factors controlling data quality. Here, we present in situ U–Pb data obtained by LA–SF–ICP–MS for a suite of widely used apatite reference materials, including NW-1, McClure Mountain, Emerald, Durango, Otter Lake, MAD, AP1 and AP2, covering an age range from ~1157 to 32 Ma. Their long-term stability and homogeneity are evaluated based on analytical results over the past five years. Using real reference materials with contrasting common Pb contents, we further assess the influence of common Pb correction using anchored initial Pb compositions with propagated uncertainties. The results show that U–Pb ages for most reference materials are consistent with ID–TIMS certified values or previously published data, with deviations ≤2%, demonstrating their long-term analytical stability. NW-1, characterized by high U concentrations and the lowest common Pb, is recommended as primary reference material, whereas AP1 and AP2 are better suited as secondary reference materials. This study confirms the robustness of the established analytical protocol, clarifies the appropriate application of different apatite reference materials, and provides a practical framework for matrix matching and multi–reference material cross-calibration in in situ apatite U–Pb geochronology.