We present a comprehensive theoretical and experimental study of the lock-in amplifier detection technique applied to two-color fluorescence thermometry. We modeled and measured the impact of different lock-in amplifier parameters (reference frequency, time constant, roll-off) and background pulse characteristics (width and steepness) on the lock-in amplifier’s background suppression, as well as modeled the impact of the lock-in amplifier parameters on temperature measurements of short duration heating events. Based on our results, we provide a guide for designing experiments using lock-in amplifiers in two-color fluorescence thermometry to obtain the best possible results.