A composition-based model is presented that predicts the effect of cetane improver (specifically 2-ethyl hexyl nitrate) on the cetane number of diesel fuels. A total of 206 different diesel fuels were considered in this work containing varying amounts of improver. The fuels were chosen to span a wide range of compositions, from highly paraffinic to highly naphthenic and, in some cases, highly aromatic fuels. Improver concentrations were varied between 0 to 3500 ppm (v/v), which exceeds the usual commercial application range of 500-1000 ppm. Detailed molecular composition of all the fuels was analyzed using a combination of GC-MS and supercritical fluid chromatographic techniques, while cetane number measurements were made using an Ignition Quality Tester (IQT). Molecular composition was correlated to the cetane number (CN) boost using a simple correlative equation, derived from phenomenological considerations. The model predicts CN boost with a standard error of +/- 0.8 CN, which is within the experimental error of the measurements.