The main aim of the present study was to estimate the carrier characteristics affecting the dissolution efficiency of griseofulvin (Gris) containing blends (BLs) using partial least squares (PLS) regression analysis. These systems were prepared at three different drug/carrier weight ratios (1/5, 1/10, and 1/20) by the solvent evaporation method, a well-established method for preparing solid dispersions (SDs). The carriers used were structurally different including polymers, a polyol, acids, bases and sugars. The BLs were characterised at the solid-state by spectroscopic (Fourier transform infrared spectroscopy), thermoanalytical (differential scanning calorimetry) and X-ray diffraction studies and their dissolution behaviours were quantified in terms of dissolution efficiencies (log DE/DE(Gris)). The correlation between the selected descriptors, including parameters for size, lipophilicity, cohesive energy density, and hydrogen bonding capacity and log DE/DE(Gris) (i.e., DE and DE(Gris) are the dissolution efficiencies of the BLs and the pure drug, respectively) was established by PLS regression analysis. Thus two models characterised by satisfactory coefficient of determination were derived. The generated equations point out that aqueous solubility, density, lipophilic/hydrophilic character, dispersive/polar forces and hydrogen bonding acceptor/donor ability of the carrier are important features for dissolution efficiency enhancement. Finally, it could be concluded that the correlations developed may be used to predict at a semiquantitative level the dissolution behaviour of BLs of other essentially neutral drugs possessing hydrogen bonding acceptor groups only.