Background and Objectives: The therapeutic effect of laser treatment for macular diseases is related to the damage to the retinal pigment epithelium (RPE) and the subsequent restoration of the defect due to RPE proliferation. In contrast to conventional laser treatment, it is possible to damage the RPE selectively and to spare the photoreceptors by using repetitive microsecond laser pulses. It was the aim of the study to investigate the influence of pulse duration and number of pulses on angiographically and ophthalmoscopically visible retinal damage thresholds in order to optimize treatment modalities. Study Design/Materials and Methods: In total, 625 laser lesions with various parameters were applied to the retina in 11 eyes of 6 Chinchilla breed rabbits using an experimental laser system (Nd:YLF at 527 nm). Pulse duration (1.7 microseconds and 200 nanoseconds) and number of pulses (100, 10, and 1 pulses) were varied at a constant repetition rate of 100 Hz. Damage thresholds were determined in terms of ophthalmoscopic and fluorescein angiographic visibility, and the therapeutic window (TW; angiographic ED 50 vs. ophthalmoscopic ED 50 ) as well as the safety range (SR; angiographic ED 84 vs. ophthalmoscopic ED 16 ) between both thresholds were calculated. Selected laser lesions were evaluated by histology. Results: Generally, the ED 50 radiant exposure for angiographic visibility decreases for shorter laser pulses and with an increase in the number of pulses. The TW for both pulse durations (1.7 microseconds and 200 nanoseconds) was wider with 100 pulses than with single pulses. The widest TW was found for 100 pulses at 200 nanoseconds pulse duration (5.9-fold above the angiographic threshold), and the smallest TW with a factor of 1.6 was found for 1.7 microseconds single pulses. In terms of SR, only irradiation with 100 pulses at 200 nanoseconds pulse duration was associated with a ratio > 2. Independently of pulse duration, histological examination of laser sites 1 hour after irradiation revealed widely intact photoreceptors, while the underlying RPE was damaged. Conclusions: Pulse duration and number of pulses have a significant influence on RPE damage thresholds and consecutively on TW and SR. Because fundus pigmentation in humans may vary intra-and interindividually by a factor of 2, a large TW and ideally also a large SR should be ensured in a clinical treatment context. In rabbits, the safety range with 200 nanoseconds pulses is higher than with the pulse duration of 1.7 microseconds currently in clinical use. These findings suggest the need for clinical pilot studies to prove whether these results can be transposed to the situation in humans.