DNA mixtures are frequently encountered in forensic casework especially in cases of sexual assault. When evidence is recovered, the sample may have come from multiple contributors in different proportions. The first part of this study examines the fidelity of contributor proportions by using the residual to analyze known mixture samples. The coefficient of determination between the expected and observed proportions was also determined and used to assess the fidelity of mixture proportions. The second part of this study involved separating major and minor contributors in a mixture by characterizing the observed proportions. Results for the 2-person mixture show that as the mass of amplified DNA decreases, the number of allele dropouts increases. Furthermore, as mass decreases, the level of variation between the expected and observed proportions increases, as determined by the residuals and the coefficients of determination. In addition, as mixture proportions become more disparate the amount of variations between the expected and observed proportions are not as great as the mass. For the 3-person mixtures, as mass decreases, the residuals increase. Also, when the coefficient of determination of the 3-person mixtures were compared to those obtained with the 2-person mixtures, it was determined that the R2 were larger for the former. This was a result of higher total amplification masses. In mixture 1:2/2:1, major and minor proportions are not distinguishable In mixture 1:4/4:1, major and minor proportions can be distinguished at 1 ng. In mixture 1:9/9:1, proportions are distinguishable at 1, and 0.5 ng. Mixtures could not be distinguished at the 0.25 ng level, despite proportion and is the result of the increase in variation with decreasing mass.