A comparison of output from three remote sensing based SVAT models is made over the SPARC2004 site in Barrax, Spain. Use is made of ASTER data obtained on the 18th of July 2004, overpass 11.008 GMT. Validation versus ground measurements showed reasonable to good results, where errors were generally within measurement accuracy. All three models reflected measured trends in net radiation, soil and sensible heat fluxes; no ground validation of latent heat flux was used. Inter-comparison of the three models learned that both net radiation and soil heat flux predictions show similar spatial variation, also here differences between the models were nearly negligible. Sensible heat flux output differed mainly between the two-source (TSEB) and the single-source (SEBS and SEBAL) schemes and could be primarily attributed to differences in surface roughness parameterization.