Objective With free tissue transfer becoming a standard procedure in reconstructive surgery, methods for detecting flap failures early are becoming more and more important. In a preceding study, indocyanine green (ICG) angiography was positively tested for detecting early stage flap malperfusion. Despite promising results, there are still some limitations to the method used concerning a lack of orientation before ICG application, as well as results depending on light and tissue geometry, distance and automatic camera settings such as integration time and gains. To overcome these, several techniques have been evaluated, including semi-quantitative fluorescence measurements within the field of observation and over fluorescence time, combining fluorescence and quasi-white-light measurements, online data acquisition and disconsolation of flap perfusion resistance. Material and methods In the preceding study, 25 patients underwent free-flap reconstruction of the upper aerodigestive tract (UADT) with flap perfusion being examined by endoscopic ICG angiography. The obtained video data was examined using deconvolution methods. The system used within this study was enhanced by adding white-light (WL) measurements to the fluorescence data as a reference light for semi-quantitative fluorescence measurements. Furthermore fluorescence kinetics was recorded, enabling online calculations such as deconvolution for comparing flap perfusion with the one of the surrounding tissue. Results As the data from the preceding study offered few prospects to study fluorescence kinetics and only some data sets were qualified for studying perfusion with deconvolution methods, a new system design was established that enables a fluorescence signal within one color channel of the camera and WL remission in the remaining two channels. A pseudo-white-light image was recalculated from the remittance signal. WL remission also was used as reference light for semi-quantitative, online fluorescence measurements. By using motion tracked ‘regions of interest’ (ROIs) on the corrected data, the recording of fluorescence kinetics was enabled without influence of lighting conditions and tissue geometry. Therefore calculations such as deconvolution of flap perfusion resistance were possible by comparing the fluorescence kinetics of the flap with the ones of the surrounding tissue. Conclusion The feasibility of endoscopic ICG fluorescence angiography in patients undergoing free-flap transfer to the UADT has been shown in a preceding study. In order to improve the method used, a new system design was evaluated that together with enhanced image processing enabled combination of the measurements from fluorescence and WL imaging. Additionally the influence of tissue geometry, lighting conditions and the camera's image processing on fluorescence signal intensity was almost entirely reduced, therefore making fluorescence measurements comparable within the observed area and over time. By tracking the fluorescence kinetics in significant ROIs and using deconvolution methods, new access to judging flap perfusion was made. Overall, the tested methods seem to not only improve the used approach but also offer new information.