Field spectroradiometer is becoming more and more important in the remote sensing area. It is used to investigate the spectral radiometric characteristics of different geographic environment, such as spectral radiance, spectral irradiance or spectral directional reflectance factor from visible to infrared wavelengths. The accuracy of the spectral measurements plays an important role in space sensor calibration and remote sensing data inversion. It is the basis of quantitative analysis in optical remote sensing. Field spectroradiometer must be calibrated before use. On the one hand, the spectral responsivity may change during the calibration process. On the other hand, the field measurement environment may be totally different from the calibration conditions. The measurement accuracy cannot be guaranteed. The experiment is designed to investigate the influence of the environmental conditions on spectral responsivity. At constant temperature and humidity, spectral line lamp, and integrating sphere lamp are used to test whether the wavelength and spectral responsivity change or not when the silicon array detector’s temperature is changed. Results show that the wavelength is nearly unchanged when the silicon array detector’s temperature rises. However, the spectral responsivity rises as the temperature rises. When the silicon array detector’s temperature rises from 28.3 to 35.2 ℃, the spectral responsivity increases by 1.8% to 7.3% from 380 to 990 nm, 3.0% from 1 000 to 1 800 nm and 1.9% from 2 000 to 2 500 nm. When the environment humidity and temperature are changed, results show that humidity only affects the spectral responsivity near the water molecules’ absorption peaks. Also, results show that the spectral responsivity has nearly one to one correspondence with detector temperature. The influence of environment temperature can be nearly characterized according to the detector temperature change. When a group of detector temperature and corresponding spectral responsivity is measured, the relationship between the detector temperature and spectral responsivity can be obtained using the least square method. The spectral responsivity at other temperature can be calculated by interpolation. The measurement data at different environmental conditions can be corrected as long as the detector temperature is measured. The spectral responsivity difference between the calculated and measured results is less than 0.2% when the detector is 35 ℃, which show that the relationship between the detector temperature and spectral responsivity can be used to solve the environment problem.