[en] The estimate of minimum detectable activity determined from counting systems is based on calculations involving background measurements. For gamma spectra systems, this estimate is derived from continuum information surrounding an anticipated or actual photo-peak and is therefore driven by the data within each spectral measurement. When the estimate and peak shape criteria are used to inject a peak into the spectrum, an observable peak is created. This concept forms the basis for testing the validity of the classical estimator when a priori nonbackground spectral information exists. A Monte Carlo simulation technique was used to generate spectra with known sample counts, average background, and spectral characteristics of the photo-peak. Two methods were used to estimate the sample counts: the classical summation of counts above background in the region of interest, and a linear regression technique proposed in this study, which uses information about the spectra shape of the photo-peak. The resulting distributions of sample count estimates from 1000 cases were tested at the 5th percentile to provide a graphical estimation of the lower limit of detection. Comparison showed that the spectra shape information provides greater sensitivity near background than commonly calculated. This implies that it is possible to report a lower minimum detectable activity when no peak is detectable. 3 refs., 8 figs