In this study, fluorescence excitation and emission matrices and multivariate curve resolution (PARAFAC) were used to detect and characterize active photosensitizers spectrally in butter. Butter samples were packed under high (air) and low oxygen (<0.05%) atmospheres and exposed to violet, green, or red light. Six photosensitizers were found: riboflavin, protoporphyrin, hematoporphyrin, a chlorophyll a-like molecule, and two unidentified tetrapyrrols. By estimation of relative concentrations, we could follow how each sensitizer was photodegraded as function of wavelength, oxygen level, and time. The degradation rate of protoporphyrin, hematoporphyrin, chlorophyll a, and one of the tetrapyrrols correlated well (0.83-0.91) with the formation of sensory measured oxidation. The results suggest that mainly type I photoreactions were responsible for the degradation of photosensitizers in both high and low oxygen atmosphere. Type II photoreactions (generation of singlet oxygen) were involved in the oxidation of butter stored in air. The study shows that PARAFAC modeling of fluorescence landscapes is an excellent tool for studying photooxidation in complex systems.