Fluorescence detectors are about three orders of magnitude more sensitive and specific making them ideal for trace analysis and complex sample matrices. However, temperature dependence of the signal is a disadvantage to fluorescence detectors. We have previously reported degradation of malondialdehyde (MDA)-thiobarbituric acid (TBA) adduct at room temperature. The present experiments tested the notion that the detector response may be blunted due to increase in temperature over time. Repeated injections of the same standard curve over 4 h found a significant effect of time on the slopes of peak area-concentration curves. When the samples were iced and injected alternating with ambient temperature standard curve samples, no differences in slopes were seen between iced and ambient temperature samples. Cooling the housing of the fluorescence lamp significantly increased the fluorescence in the samples. Fluorescence increased 2.5% (95% confidence interval, 1.5-3.6%) for each 1 degrees C fall in temperature. MDA-TBA adduct remained stable at room temperature. Since the fluorescence signal is temperature sensitive, letting the detector warm up for 2 h to obtain a steady temperature is more likely to give reproducible results compared to a detector that has not warmed up. These results have implications for other applications utilizing fluorescence detectors.