The objective of this research is to determine whether there are significant differences in the (18)F-FDG produced by either the phosphate or the citrate buffer cassettes in the FASTlab synthesizer. METHODS Forty batches of (18)F-FDG were produced with each cassette and analyzed retrospectively. The analysis consisted of determining the mean radiochemical yield (RCY)-uncorrected and corrected for decay-radiochemical purity (RCP), pH, and residual solvent content (ethanol and acetonitrile). An independent t test (alpha error [α], 0.05) was performed to determine whether the differences were statistically significant. RESULTS The mean decay-corrected RCYs for (18)F-FDG produced by phosphate and citrate cassettes were 82.9% ± 17.4% and 79.2% ± 5.0%, respectively. The uncorrected RCY was 57.5% ± 16.7% for phosphate- and 58.8% ± 6.0% for citrate-buffered (18)F-FDG, leading to a difference of 4.4% and P value of 0.11 for corrected RCY and a difference of 2.2% and P value of 0.32 for uncorrected RCY. Thus, the RCY differences are neither statistically nor clinically significant. The mean RCPs were 99.4% ± 0.2% for the phosphate-buffered (18)F-FDG and 99.0% ± 1.1% for the citrate-buffered (18)F-FDG. There was a 0.5% difference and a P value of 0.021, meaning that the difference was statistically significant. The average pHs for (18)F-FDG produced by phosphate and citrate buffer cassettes were 5.9 ± 0.1 and 5.3 ± 0.2, respectively, resulting in a 9.6% difference and a P value close to zero (2.6 × 10(-19))-a statistically significant difference. The difference between ethanol content was also dramatic. Phosphate-buffered (18)F-FDG contained 0.08% ± 0.02% ethanol, whereas the citrate-buffered (18)F-FDG contained 0.20% ± 0.07%. No difference was found in the acetonitrile content of the 2 cassettes. CONCLUSIONS The differences in yield between cassettes are due to statistical variability. The results confirm our hypothesis that there is no significant difference in RCY. The differences seen in the statistically significant data (those with a P value > 0.05) turn out to be insignificant in a real-world setting because all values fell within the limits set by the United States Pharmacopeia and Food and Drug Administration. Therefore, determining which cassette to use is a matter of the preference of the institution.