Various factors affecting the random and systematic error in the repeatability and reproducibility of thermoluminescent 6 LiF:Mg,Ti dosimeters were investigated. These characteristics were examined by irradiating them at different doses in the range 3.8 µG to 433.94 mGy from a Cs-137 gamma source. Values of the element correction factors (ECFs) were obtained by exposing the chips to a dose of 433.93 mGy five times. This investigation denotes that more correction will be obtained by calculating the error due to non-subtracting zero dose reading from the delivered dose, which decreases the coefficient of variation as a percentage of the mean from the range of 2.4%-18.6% to only 0.7%-16.9%. Using the individual sensitivity factor and repeating five calibrations decreases the coefficient of variation as a percentage of the mean to only 0.54%. The values of the Variability Index for System (SVI), Reader (RVI) and Detector (DVI) are 1.46 %, 1.40% and 0.34% respectively. Also, it is clear from the results that the best repeatability and reproducibility obtained by using the element correction factor was associated with each dosimeter rather than carefully selecting the individual dosimeters. Finally, from the results, it can be noted that at low doses, the value of relative coefficient of variation becomes very large, while at high doses, it becomes basically constant and of the order of a few percentage.