Abstract Introduction It is important to identify all circulating metabolites, including free fluoride, for accurate pharmacokinetic modeling of 18 F-labeled radiotracers. We sought to determine the most efficient method to detect and quantify low levels of free 18 Ffluoride in biological samples. Methods Low levels of 18 Ffluoride were analyzed using two methods: (A) an ion-exchange cartridge and gamma counting, and (B) radio-HPLC, to compare the detection limits of these two analytical methods. Twenty microliters of 18 Ffluoride solution was loaded onto an ion-exchange cartridge, then eluted with 20% MeCN/water (5 ml) and radioactivity trapped in the cartridge counted on a gamma counter. 18 FFluoride was also determined in plasma and urine from mice injected with 18 F-labeled thymidine analogues using Method A. Results The detection sensitivity of Method A was 9.4-fold higher than that of Method B (0.075±0.004 vs. 0.71±0.02 nCi). With Method A, 18 Ffluoride was determined in plasma for 18 FFLT, 18 FFMAU, 18 FFEAU and N3 -18 FFPrT as 1.4±0.31% ( n =4), 0.17±0.49% ( n =3), 4.88±1.62% ( n =3) and 12.94±0.48% ( n =4), respectively. The amount of 18 Ffluoride determined in the urine was 11.49±1.60% ( n =4) from 18 FFLT, 5.36±2.34% ( n =3) from 18 FFMAU, 13.57±1.96% ( n =3) from 18 FFEAU and 11.19±1.98% ( n =4) from N3 -18 FFPrT. Conclusion Low levels of 18 Ffluoride in biological samples can be detected and quantified using an ion-exchange cartridge and gamma counting. This methodology is simple, accurate and superior to the standard use of radio-HPLC on a C18 column for metabolite analysis, and it should be useful in pharmacokinetic modeling for animal imaging studies using an 18 F-labeled radiotracer and PET.