We study the effect of the coronal background in the determination of the diameter of EUV loops, and we analyze the suitability of the procedure followed in a previous paper (Lopez Fuentes, Klimchuk & Demoulin 2006) for characterizing their expansion
properties. For the analysis we create different synthetic loops and we place them on real backgrounds from data obtained with the Transition Region and Coronal Explorer (textit{TRACE}). We apply to these loops the same procedure followed in our previous works, and we compare the results with real loop observations. We demonstrate that the procedure allows us to distinguish constant width loops from loops that expand appreciably with height, as predicted by simple force-free field models. This holds even for loops near the resolution limit. The procedure can easily determine when loops are below resolution limit and therefore not reliably measured. We find that small-scale variations in the measured loop width are likely due to imperfections in the background subtraction. The greatest errors occur in especially narrow loops and in places where the background is especially bright relative to the loop. We stress, however, that these effects do not impact the ability to measure large-scale variations. The result that observed loops do not expand systematically with height is robust.
