Electrical resistance and low-frequency noise of Mn-Co-Ni type infrared detectors were measured during space-simulated environment experiments. Two groups of samples were irradiated up to 150 krad(si) accumulated dosage with a cobalt-60 source at dose rate of 10 rad(si)/s and 0.1 rad(Si)/s), respectively. The results show that 1/f noise increased significantly at dose rate of 0.1rad(Si)/s, but almost unchanged at dose rate of 10 rad(Si)/s. Three thermal stresses, namely, short-term thermal burn-in without bias (40℃, hold for 4 h), extended thermal brun-in with DC bias (bias voltage of ±15 V, 40℃, hold for 600 h) and thermovacuum cycling (－40℃ to +40℃, 1℃/min, hold for 1 h, 20 cyclings), were successively applied to a third group of samples, the changes of electrical resistance of samples have similar behavior; however, the degradation of 1/f noise shows different trends, and all malfunctions present abrupt increment. Theoretical analysis shows that the degradations of 1/f noise during the short-term thermal burn-in and extended thermal burn-in with DC bias may be ascribed to the defects within the thermistor bulk, and the degradations of 1/f noise during thermal cycling may be ascribed to the defects at the contacts. These results suggest that noise figure is a sensitive parameters for describing the low-frequency noise degeneration of infrared detectors, and short-term thermal burn-in and thermal cycling are effective testing methods for discriminating 1/f noise of infrared detectors.