In recent years there has been a great deal of interest in the use of fluorescent particles to quantify ocular blood flow. The extraction and analysis of data from the markers, however, are tedious and subjective. This has limited the utility of particle-tracking techniques, and has suggested the need for a means of automating these time-consuming tasks. The objectives of this research are: to develop a fluorescent particle suitable for imaging the ocular fundus; to develop and validate computer software to automate the extraction of data from particle-tracking sessions; to evaluate the reproducibility of the software; to use the software to map hemodynamic parameters in the normal rabbit fundus; and to evaluate the software as a tool for detecting pathological changes in retinal blood flow In developing fluorescent microsphere imaging, we evaluated and optimized single- and multiple-dye polystyrene microspheres as tools for imaging retinal and choroidal blood flow. These particles are injected systemically and excited transcorneally using the lasers of the scanning laser ophthalmoscope. The ophthalmoscope then creates a video image of the particle emissions. By tailoring the excitation and emission spectra of the particles, we can selectively image blood flow at various depths and in various tissues of the ocular fundus We developed a series of computer programs with which videotapes of these imaging sessions can be automatically evaluated. After detecting the positions of particles, we generated maps of relative volumetric blood flow in the entire field of view of the ophthalmoscope. Further, we developed routines for quantifying particle motions and used the information to map absolute blood velocities in the fundus. We combined this information with our maps of vessel diameters to measure absolute blood flow rates in large vessel segments; this provided a scale with which we mapped absolute flow rates across the entire ocular fundus We then tested these techniques by imaging particles in the eyes of four rabbits. Each rabbit was evaluated three times, including one session in which we induced abnormal flow. In a series of ANOVA tests, we demonstrated the ability of the software to distinguish normal from pathological ocular blood flows