A common method of optimizing coolant performance in gas turbine engines is through the use of shaped film-cooling holes. Despite widespread use of shaped holes, existing correlations for predicting performance are limited to narrow ranges of parameters. This study extends the prediction capability for shaped holes through the development of a physics-based empirical correlation for predicting laterally-averaged film-cooling effectiveness on a flat plate downstream of a row of shaped film-cooling holes. Existing data was used to determine the physical relationship between film-cooling effectiveness and several parameters, including; blowing ratio, hole coverage ratio, area ratio, and hole spacing. Those relationships were then incorporated into the skeleton form of an empirical correlation, using results from the literature to determine coefficients for the correlation. Predictions from the current correlation, as well as existing shaped hole correlations and a cylindrical hole correlation were compared to the existing experimental data. Results show that the current physics-based correlation yields a significant improvement in predictive capability, by expanding the valid parameter range and improving agreement with experimental data. Particularly significant is the inclusion of higher blowing ratio conditions (up to M = 2.5) into the current correlation, whereas the existing correlations worked adequately only at lower blowing ratios (M~ 0.5).