A data set obtained on 2003 January 17 with the Coronal Diagnostic Spectrometer (CDS) shows two loops sitting side by side on the solar disk. These loops are oriented along the CDS slit, so all pixels in each loop were observed simultaneously. So, although the instrument has a relatively slow time cadence, changes as a function of time that may occur during the CDS raster buildup will not affect the loop temperature results. Differential emission measure (DEM) analysis using a forward-folding technique shows different results for the two loops. For the first loop, the intensities of the lines that remain after background subtraction are well fit with a DEM curve that collapses to a single spike. In other words, the loop plasma at this location is isothermal. This analysis is confirmed with an emission measure loci method and agrees with the results obtained recently by other authors that show that the moderate spatial resolution of CDS can detect isothermal structures. For the second loop, the background-subtracted line intensities require a broad DEM, not consistent with isothermal plasma. This conclusion is confirmed with an automatic-inversion DEM method. In this Letter, we specifically address some of the concerns raised about CDS temperature analysis: the slow CDS temporal resolution, the moderate CDS spatial resolution, the inherent smoothing associated with DEM inversion, and line-of-sight effects on the DEM distribution.