Several individual clinical and preclinical studies have shown the possibility of evaluating tumor hypoxia by using noninvasive positron emission tomography (PET). The current study compared 3 hypoxia PET tracers frequently used in the clinic, 18FFMISO, 18FFAZA, and 18FHX4, in a preclinical tumor model. Tracer uptake was evaluated for the optimal time point for imaging, tumor-to-blood ratios (TBR), spatial reproducibility, and sensitivity to oxygen modification. PET/computed tomography (CT) images of rhabdomyosarcoma R1-bearing WAG/Rij rats were acquired at multiple time points post injection (p.i.) with one of the hypoxia tracers. TBR values were calculated, and reproducibility was investigated by voxel-to-voxel analysis, represented as correlation coefficients (R) or Dice similarity coefficient of the high-uptake volume. Tumor oxygen modifications were induced by exposure to either carbogen/nicotinamide treatment or 7% oxygen breathing. TBR was stabilized and maximal at 2 hours p.i. for 18FFAZA (4.0 ± 0.5) and at 3 hours p.i. for 18FHX4 (7.2 ± 0.7), whereas 18FFMISO showed a constant increasing TBR (9.0 ± 0.8 at 6 hours p.i.). High spatial reproducibility was observed by voxel-to-voxel comparisons and Dice similarity coefficient calculations on the 30% highest uptake volume for both 18FFMISO (R = 0.86; Dice coefficient = 0.76) and 18FHX4 (R = 0.76; Dice coefficient = 0.70), whereas 18FFAZA was less reproducible (R = 0.52; Dice coefficient = 0.49). Modifying the hypoxic fraction resulted in enhanced mean standardized uptake values for both 18FHX4 and 18FFAZA upon 7% oxygen breathing. Only 18FFMISO uptake was found to be reversible upon exposure to nicotinamide and carbogen. This study indicates that each tracer has its own strengths and, depending on the question to be answered, a different tracer can be put forward.