The objective of this study is to compare the performance of the DIAGNOdent laser fluorescence (LF) Pen to conventional periodontal probing for detection of subgingival calculus under defined laboratory conditions. Extracted teeth with various levels of subgingival deposits of calculus were mounted anatomically in stone casts, and an impression material was used to replicate periodontal soft tissues. The casts were examined for the presence of subgingival calculus at eight surfaces per tooth (240 sites) using LF and a periodontal probe. Sites were rescored after 1 and 3 weeks. Direct imaging of the root surfaces under magnification was the gold standard. As a result, for an experienced operator, LF was more accurate than tactile assessment (across all sites, 84.0% vs. 59.8%). The performance difference was greater for multi‐rooted teeth (85.8% vs. 56.9%) than single‐rooted teeth (77.2% vs. 66.7%). The performance of LF in this laboratory trial was influenced strongly by clinician skill and experience. When used by an experienced operator, LF was more sensitive (75.1% vs. 69.2%), specific (92.6% vs. 86.3%), and accurate (84% vs. 77.9%) than for an inexperienced operator. In conclusion, under the defined laboratory conditions used, LF had better performance than tactile examination, particularly for multi‐rooted teeth. This method may have value clinically as an adjunct for detecting subgingival deposits of calculus in clinical practice. The usefulness of the method improves with operator experience.