The fluorescence intensity of fluorescent dyes typically employed in the dual-emission laser-induced fluorescence (DELIF) method gradually degrades as the excitation time increases, and the degradation rate depends on the type of fluorescent dye used. Therefore, the DELIF method is unsuitable for long-term temperature measurements. In this study, we focused on the fluorescence intensity ratio of a single fluorescent dye at two fluorescence wavelengths and developed a DELIF method for long-term temperature measurements based on this ratio. The fluorescence intensity characteristics of Fluorescein disodium and Rhodamine B, which are typically used in the DELIF method, in the temperature range of 10–60 °C were comprehensively investigated using two high-speed monochrome complementary metal-oxide semiconductor cameras and narrow bandpass filters. Interestingly, the ratio of the fluorescence intensity of each fluorescent dye at the peak emission wavelength of the fluorescence spectrum, λ, to the fluorescence intensity at wavelengths very close to the peak wavelength (λ ± 10 nm) was highly sensitive to temperature variations but not excitation time. Particularly, when Rhodamine B was used, the selection of the fluorescence intensity ratios at a wavelength combination of 589 and 600 nm enabled highly accurate temperature measurements with a temperature resolution of ≤0.042 °C. Moreover, the fluorescence intensity ratio varied negligibly throughout the excitation time of 180 min, with a measurement uncertainty (95% confidence interval) of 0.045 °C at 20 °C. The results demonstrate that the proposed DELIF method enables highly accurate long-term temperature measurements.