Allosteric modulators of the metabotropic glutamate receptor subtype 5 (mGlu5) have exciting potential as therapeutic agents for multiple brain disorders. Translational studies with mGlu5 modulators have relied on mGlu5 allosteric site positron emission tomography (PET) radioligands to assess receptor occupancy in the brain. However, recent structural and modeling studies suggest that closely related mGlu5 allosteric modulators can bind to overlapping but not identical sites, which could complicate interpretation of in vivo occupancy data, even when PET ligands and drug leads are developed from the same chemical scaffold. We now report that systemic administration of the novel mGlu5 positive allosteric modulator VU0092273 displaced the structurally related mGlu5 PET ligand, [(18)F]FPEB, with measures of in vivo occupancy that closely aligned with its in vivo efficacy. In contrast, a close analog of VU0092273 and [(18)F]FPEB, VU0360172, provided robust efficacy in rodent models in the absence of detectable occupancy. Furthermore, a structurally unrelated mGlu5 negative allosteric modulator, VU0409106, displayed measures of in vivo occupancy that correlated well with behavioral effects, despite the fact that VU0409106 is structurally unrelated to [(18)F]FPEB. Interestingly, all three compounds inhibit radioligand binding to the prototypical MPEP/FPEB allosteric site in vitro. However, VU0092273 and VU0409106 bind to this site in a fully competitive manner, whereas the interaction of VU0360172 is noncompetitive. Thus, while close structural similarity between PET ligands and drug leads does not circumvent issues associated with differential binding to a given target, detailed molecular pharmacology analysis accurately predicts utility of ligand pairs for in vivo occupancy studies.