The white and additive noise generated by an R-SQUID noise thermometer whose temperature was varied from 6.3 mK to 0.738 K is measured. Room temperature simulations of the effect of white and additive noise on the circuit used to measure the noise for the R-SQUID are conducted. It is shown that the measured noise in all cases was fit to within the 0.1% statistical measurement imprecision by a model for the R-SQUID based on frequency modulation. This conformity is sufficiently good that any deficiencies in either do not lead to systematic inaccuracies in the noise temperature scale which exceed 0.1%. Comparisons of the R-SQUID with several other thermometers indicate that the overall agreement is at worst 0.2%.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>