Fluorescence lifetime imaging ophthalmoscopy (FLIO) has developed as a new diagnostic tool in ophthalmology. FLIO measurements are taken from 30° retinal fields in two spectral channels (short spectral channel (SSC): 498–560 nm, long spectral channel (LSC): 560–720 nm). Because of the layered structure of the eye, the detected signal is an interaction of the fluorescence decay of the anterior part and of the fundus. By comparing FLIO measurements before and after cataract surgery, the impact of the natural lens was proven, despite the application of a confocal laser scanning (cSLO) technique. The goal of this work was to determine the best algorithmic solution to isolate the sole fundus fluorescence lifetime from the measured signal, suppressing artifacts from the natural lens. Three principles based on a tri-exponential model were investigated: a tailfit, a layer-based approach with a temporally shifted component, and the inclusion of a separately measured fluorescence decay of the natural lens. The mean fluorescence lifetime τ_m,12 is calculated using only the shortest and the intermediate exponential component. τ_m,all is calculated using all three exponential components. The results of tri-exponential tailfit after cataract surgery were considered as a reference, because the implanted artificial lens can be assumed as non-fluorescent. In SSC, the best accordance of τ_m,all of the reference was determined with τ_m,12 of the tailfit before surgery. If high-quality natural lens measurements are available, the correspondence of τ_m,12 is best with τ_m,all of the reference. In LSC, there is a good accordance for all models between τ_m,12 before and after surgery. To study the pure fundus fluorescence decay in eyes with natural lenses, we advise to utilize fluorescence lifetime τ_m,12 of a triple-exponential tailfit, as it corresponds well with the mean fluorescence lifetime τ_m,all of eyes with fluorescence-less artificial intraocular lenses.