Apatite provides a powerful means to determine the Sr isotopic composition (87Sr/86Sr) of ancient crust as it contains enough Sr to be measured accurately and precisely while typically having little post‐crystallisation radiogenic ingrowth due to its low Rb. Although utilising LG‐SIMS is beneficial for small‐volume targets, for example, inclusions, it is technically challenging to measure Sr isotopes in apatite with good accuracy and precision by SIMS, mainly due to molecular isobars interfering on Sr isotopes that have not previously been thoroughly assessed. This study evaluates all theoretically possible interferences affecting Sr isotope analysis of apatite by SIMS and develops a principle protocol for analysis and data processing that can be applied across a wide range of Sr concentration, even lower than 100 µg g−1. With careful control of the peak shape and applying precise mass centring at an intermediate mass resolution (ca. 4500), only interferences from the apatite matrix‐produced Ca dimers and 40Ca31P16O (present as a low mass tail) were found to be significant for the low trace element reference apatite used in this study, both of which can be precisely measured and corrected for. Despite this comprehensive approach, the Ca2 correction remained inadequate, possibly due to as yet unidentified doubly charged molecules interfering at the reference mass 82 peak (40Ca42Ca). While this can be reduced by employing an energy offset of at least −25 eV, precision significantly worsens in low Sr apatite (ca. 0.0024 in 1SE compared with ca. 0.0010 without offset for < 100 Sr µg g−1 apatite). Use of an empirical correction for the residual mass 82 interference is therefore preferred to an energy offset analytical approach, especially for low Sr apatite. Lastly, a slightly positive offset (ca. +0.0007) of unknown origin in 87Sr/86Sr was consistently observed across sessions with different analytical conditions, which may also be empirically corrected. The final corrected 87Sr/86Sr of the reference materials show good accuracy and precision with repeatability from ca. 0.0010 to 0.0002 in 1SE obtained for Sr contents from < 100 to 1500 µg g−1.