We develop a diagnostic method for the coronal heating mechanism in active region loops. Observational constraints on coronal heating models have been previously derived from statistical analysis of power laws, using integrated EUV emission from many active regions by Fludra and Ireland (2003, 2008). Fludra and Warren (2010) analysed fully resolved images in a spectral line of OV 63.0 nm, identified a dominant variable component of the transition region emission and a steady basal heating, and derived the dependence of the basal heating rate on the photospheric magnetic flux density. In this study, we compare models of single coronal loops with EUV observations, assessing whether observations of individual loops are capable of providing constraints on the heating mechanism. We derive the coronal magnetic field in an active region using an NLFF extrapolation and model the plasma in selected loops using a 1D hydrostatic code, applying a heating rate as a function of magnetic field strength along the loop. We calculate the EUV emission along the loops in SDO/AIA 171 and 335 bands, and in pure spectral lines of Fe IX 17.1 nm and Fe XVI 33.5 nm. We use different distributions of the heating function: from concentrated near the loop-top, uniform, to concentrated near the footpoints, and investigate their effect on the modelled EUV intensities and the existence of stable solutions. We find a diagnostics based on the dependence of the total loop intensity on the shape of the heating function and discuss its range of applicability for different loops.