Electrospun (ES) non-woven matrices are regarded as promising platforms for the development of miniaturized sensing systems with improved detection capacity. Their high specific surface area and void-to-volume ratio are expected to promote higher and faster interaction of the sensing molecular probes with the target stimuli increasing sensor sensitivity and response time. However, the poor light transparency of ES non-woven mats appears as the main limiting effect regarding their use as optical sensor platforms being important to determine to what extent it affects the sensor prediction accuracy. This work addresses this question providing a comparative analysis of the performance of flat cast and ES non-woven cellulose acetate luminescent platforms loaded with a pH sensitive dye-pair, i.e. fluorescein isothiocyanate (FITC) and rhodamine 6G (R6G). This study follows a comprehensive approach aiming at clarifying the effect of the platform morphology on the sensitivity of their spectral properties to pH and to understand about the advantages of using dual dye systems for pH detection. The presence of R6G was found to improve remarkably the sensitivity of these ES matrices extending the analytical capacity of the probes to the alkaline range. The emission of ES matrices showed stronger sensitivity to pH. However, pH prediction accuracy was found to depend crucially on a synergistic effect from the platform morphology and the signal analysis methodology. ES non-woven matrices allows for accurate pH prediction, characterized by determination errors < 10% in the whole pH < 10, by exponential analysis of the dye-pair emission at λExc of 450 nm and λEm of 519 nm. Furthermore, it shows a strong reduction of the determination errors, at extremely acidic conditions, resulting in values comparable to that obtained by analysis of the emission signal from flat cast platforms with more complex Förster Resonance Energy Transfer (FRET) methodologies.