Two-photon laser scanning microscopy (TPLSM) of endogenous reduced nicotinamide adenine dinucleotide (NAD(P)H) provides important information regarding the cellular metabolic state. When imaging the punctate mitochondrial fluorescence originating from NAD(P)H in a rat basophilic leukemia (RBL) cell at low laser powers, no morphological changes are evident, and photobleaching is not observed when many images are taken. At higher powers, mitochondrial NAD(P)H fluorescence bleaches rapidly. To assess the limitations of this technique and to quantify the extent of photodamage, we have measured the effect of TPLSM on DNA synthesis. Although previous reports have indicated a threshold power for "safe" two-photon imaging, we find the laser power to be an insufficient indicator of photodamage. A more meaningful metric is a two-photon-absorbed dose that is proportional to the number of absorbed photon pairs. A temporary reduction of DNA synthesis in RBL cells occurs whenever a threshold dose of approximately 2 x 10(53) photon2 cm-4 s-1 is exceeded. This threshold is independent of laser intensity when imaging with average powers ranging from 5 to 17 mW at 740 nm. Beyond this threshold, the extent of the reduction is intensity dependent. DNA synthesis returns to control levels after a recovery period of several hours.