We have previously demonstrated the detection of reversible and irreversible changes on MR images oflaser energy deposition and tissue heating and cooling1. It is possible to monitor and control energy deposition during interstitial laser therapy. This presentation describes some first steps toward optimizing the power and total energy deposited in various tissues in vivo, by analyzing the irreversible tissue changes and their spatial distribution as revealed by spin echo imaging. We used various power settings of an Nd.YAG laser delivered by a fiber optic inserted into several tissues (brain, muscle, liver) of anesthetized rats and rabbits. MR imaging was performed at 1.9 T. Photothermally-produced lesions were seen on both T1- and Ta-weighted images. The overall size of the lesions correlated with the magnitude of the energy applied. The MR image appearance depended not only on the laser energy but also on the way it was delivered, on the type of tissue, and the MR pulse sequence applied. While Ti-weighted images adequately demonstrated an area of tissue destruction, T2- weighted images showed a more heterogeneous and more extensive lesion which could be better correlated with the complex histological representation of these lesions. Typically, when rabbit brain, liver, and muscle had been exposed to laser power of 2.5 Watts for a range of 55 to 120 seconds, depending on the tissue, a central area of signal void was surrounded by an inner hypointensity and an outer hyperintensity on T2-weighted images. The 3D extent of the lesions was well demonstrated on multislice images, providing correlation of the affected volumes seen on MRI with volumes seen in histological or histochemical preparations. We are developing an analytical model of laser heating and its effect on MR images to assess whether heating during imaging will produce unacceptable artifacts during surgery. The effect of heating is modeled as a change in magnetization during image acquisition. The region in which the change occurs is blurred by the Fourier transform of the change in magnetization as a function of time. Thus, blurring is minimized when changes occur slowly, compared to image acquisition times. We conclude that MRI can demonstrate the 3D extent of the lesions induced by lasers and can be used to investigate and optimize the control of induced tissue change within the affected volume.