We read with interest the article by Sachs and Bilfinger (August 2005)1 on the impact of positron emission tomography (PET) on clinical decision making in an academic lung cancer center and wish to comment regarding the consistency of the data and also on the generalizability of the findings. In calculating the false-positive rate of PET for the primary site, the authors grouped together both positive and indeterminate results (ie, any nonzero standardized uptake value [SUV]). This grouping was undertaken to “subject PET scan to the most rigorous standard of accuracy.” However, to calculate the positive predictive value (PPV) of PET, they considered only positive results (ie, >2.5 SUV), thus quoting a PPV of 87.3%. We suggest that this estimate of PPV is confusing, because if one calculates the PPV using the provided false-positive rate, the PPV of PET is 63.3%. Recognizing that these different estimates reflect different inclusion criteria, we submit that the lower estimates for the false-positive rate and PPV warrant mention.