In strong sunglint areas the brightness temperatures in channel 3 (3.7 mu m) and channel 4 (11 mu m) of the Advanced Very High Resolution Radiometer (AVHRR) have been found to differ by more than 27 degK. In the worst case, where there is a large amount of water vapour, as in the tropics, the difference in atmospheric effects in these two channels is only about 3 degK. Therefore, most of the observed temperature difference must be due to specular reflection of the solar radiation. Although detection of intense sunglint has been reported by others their analysis was in terms of digital numbers which cannot reveal if the detector is saturated. Using brightness temperatures rather than digital numbers, it is shown that the magnitude of sunglint contribution cannot be fully quantified because the 3.7 mu m sensor is saturated. There are two independent lines of evidence which show the saturation of the detector: ( i ) In the absence of sunglint there is a periodic noise in channel 3 of the NOAA-7 AVHRR which disappears in the strong sunglint zone and, ( ii ) the sunglint contributions in channel 1 (0.63 mu m) and channel 2 (0.93 mu m) of the AVHRR change in those areas where the sunglint in the 3.7 mu m channel remains constant. The observed saturation temperature is about 319 degK, whereas that calculated from the inflight calibration parameters is about 323degK. Assuming that the inflight calibration information is reliable, this discrepancy in the saturation temperature is explained by proving that the full quantization range of 0 to 1023 is not utilized. Therefore, saturation of the detector cannot be ascertained from the values of digital numbers.