This paper has investigated the effect of operating parameters such as ambient temperature and applied voltage on the dead time of a thin-walled Geiger-Muller (GM) counter using non-paralyzing model and two-source method. Experimental studies have been conducted using 137Cs and 90Sr sources at voltages ranging from 600 to 800 V and in the temperature range of -27-70 0C. The results of the investigations for applied voltage indicate that the dead time behavior in terms of voltage can be classified into three distinct regions. In region I (low voltages), the dead time decreases with increasing voltage, in region II (voltage close to the operating voltage), the dead time is almost constant. The dead time in region III (voltages above 740V) increases slowly and linearly with increasing voltage. The variation in the dead time in the region I is greater than region III. Region II with the minimum dead time and minimum variation of the operating voltage is the best operating region. Studies show that the variation of dead time and the range of the dead time plateau (District II) for 137Cs and 90Sr sources is different. The dead time was using 137Cs source was obtained between 58 and 78 ms and using 90Sr source between 79 and 130 ms. In general, the variations of dead time versus voltage for each of Regions I, II and III for 90Sr source are lower than 137Cs source. Experimental results also show that the dead time increases with increasing temperature.