Noninvasive functional imaging of tumors can provide valuable early-response biomarkers, in particular, for targeted chemotherapy. Using various experimental tumor models, we have investigated the ability of positron emission tomography (PET) measurements of 2-deoxy-2-[(18)F]fluoro-glucose (FDG) and 3'-deoxy-3'-[(18)F]fluorothymidine (FLT) to detect response to the allosteric mammalian target of rapamycin (mTOR) inhibitor everolimus. Tumor models were declared sensitive (murine melanoma B16/BL6 and human lung H596) or relatively insensitive (human colon HCT116 and cervical KB31), according to the IC(50) values (concentration inhibiting cell growth by 50%) for inhibition of proliferation in vitro (<10 nM and >1 microM, respectively). Everolimus strongly inhibited growth of the sensitive models in vivo but also significantly inhibited growth of the insensitive models, an effect attributable to its known anti-angiogenic/vascular properties. However, although tumor FDG and FLT uptake was significantly reduced in the sensitive models, it was not affected in the insensitive models, suggesting that endothelial-directed effects could not be detected by these PET tracers. Consistent with this hypothesis, in a well-vascularized orthotopic rat mammary tumor model, other antiangiogenic agents also failed to affect FDG uptake, despite inhibiting tumor growth. In contrast, the cytotoxic patupilone, a microtubule stabilizer, blocked tumor growth, and markedly reduced FDG uptake. These results suggest that FDG/FLT-PET may not be a suitable method for early markers of response to antiangiogenic agents and mTOR inhibitors in which anti-angiogenic/vascular effects predominate because the method could provide false-negative responses. These conclusions warrant clinical testing.