On conventional PET/CT, and under physiological conditions, the volume of the pituitary gland (PG) is small, and its metabolic activity is commonly comparable to the surrounding background level in ¹⁸ F-FDG imaging. We compared the physiological ¹⁸ F-FDG uptake of the PG in patients imaged with digital PET (dPET) and with conventional PET (cPET). Additionally, we performed phantom experiments to characterize signal recovery and detectability of small structures. We retrospectively included 10 dPET and 10 cPET patients and measured PG SUVmax, SUVmean and SUVratio (using cerebellum as reference). We imaged a modified NEMA/IEC phantom with both dPET and cPET (background activity 5 kBq/mL, and 3× and 5× higher concentrations in ∅2-20-mm spherical inserts). Mean recovery coefficients (RCmean) and signal-difference-to-noise-ratio (SDNR) were computed to assess lesion detectability. Patients imaged with dPET presented higher PG SUVmax and SUVratio (SUVR) compared to patients imaged with cPET (4.7 ± 2.05 vs. 2.9 ± 0.64, p = 0.004; and 0.62 ± 0.25 vs 0.39 ± 0.09, p = 0.029, respectively), while there was no difference for SUVmean (2.7 ± 1.32 vs 2.1 ± 0.44, p = 0.39). Thus, with a SUV readout scale of 0-5 g/mL, normal PG appeared abnormally hot with dPET, but not with cPET. Phantom evidenced higher RCmean in dPET compared to cPET. For both 3x and 5x measurements, lesion detectability according to size was systematically superior with dPET. In conclusion, patients imaged with dPET presented higher ¹⁸ F-FDG physiological uptake of the PG as compared to patients imaged with cPET. These findings were supported by phantom experiments demonstrating superior signal recovery and small region detectability with dPET. Awareness of this new "higher" SUV of the normal ¹⁸ F-FDG uptake of the PG is important to avoid potential pitfalls in image interpretation, notably in oncologic patients treated with immunotherapy, who are at increased risk to develop hypophysitis.