Xenotime is an ideal mineral for U-Th-Pb isotopic dating because of its relatively high U and Th contents, but typically low concentration of common Pb. These characteristics, and the fact that it is widespread throughout various types of rocks, suggest that the U-Th-Pb dating of xenotime has broad applications. Studies of U-Pb dating on xenotime by ion microprobe (such as SHRIMP) have increased in recent years, whereas studies by laser ablation (LA)-ICP-MS are still rare. In this study, we developed a technique for U-Pb dating of xenotime using the 193 nm ArF laser-ablation system and Agilent 7500a Q-ICP-MS. To evaluate the reliability of our method, a xenotime standard, BS-1, was analyzed and calibrated against another xenotime standard, MG-1. The weighted mean 206 Pb/ 238 U ages of 510.1 ± 5.2 Ma (2 n = 21), 509.8 ± 4.3 Ma (2 n = 21) and 510.0 ± 4.6 Ma (2n = 21) were obtained using beam diameters of 16, 24 and 32 m, respectively. These ages are identical to those determined by ID-TIMS method (weighted mean 206 Pb/ 238 U age of 508.8 ± 1.4 Ma), which supports the reliability of our LA-ICP-MS method. We also analyzed xenotimes in leucogranites from South Tibet and granites from Xihuashan in southern China, and obtained accurate and precise ages. Nevertheless, we observed systematic differences in Pb/U fractionation among xenotime, monazite and zircon. The matrix-effect resulted in either under-correction or over-correction of fractionation, and thus led to inaccurate ages. Thus, a matrix-matched material is required for U-Pb dating of xenotime by LA-ICP-MS.