The metabolic response of normal rat brain to photodynamic therapy (PDT) was studied over a 1 week interval using in vivo 31P-NMR spectroscopy. Rats injected with 12.5 mg/kg Photofrin II were submitted to brain photoactivation 48 h after drug administration with either 140 or 70 J/cm2 light (630 +/- 1 nm) from an Argon dye laser. Control studies, animals not given drug or light, animals submitted only to brain illumination without drug, and animals given drug but no light, were also performed. The data revealed a transient metabolic degradation; a decrease in the ratio of beta-nucleotriphosphate to inorganic phosphate (P less than 0.001) at 24 h after PDT treatment was followed by a return to pretreatment spectral values. Brain tissue alkalosis was also noted, with significant (P less than 0.05) differences in brain tissue pH detected at 72 h post treatment between 70 J/cm2 PDT vs control studies and at 1 week post treatment between 140 J/cm2 vs 70 J/cm2, 140 J/cm2 vs no light-no drug and 140 J/cm2 vs drug only. The data suggest that there is no difinitive metabolic marker from 31P-NMR spectroscopy that can identify necrotic brain tissue caused by PDT. Phosphorus-31 NMR data are also presented which suggest that PDT damage to brain is not solely the result of microvascular occlusion causing ischemic necrosis.