Polygenic risk scores (PRS) quantify the genetic contribution of an individual’s genotype to a trait, e.g. disease or phenotype. PRS can be used to group subjects into different risk strata and can thus be treated as predictors in clinical and epidemiological studies to help us better understand such traits. However, PRS methods typically focus on a single trait at a time, ignoring the potential simultaneous influence of a gene on multiple traits. To address this limitation in the existing methods, we propose a model that incorporates a genetic correlation matrix among traits into the cost function of a penalized regression framework with the objective to improve the predictive ability of multi-trait PRS models. The proposed method is evaluated against alternative approaches when both a single trait and multiple traits are of interest via comprehensive simulation studies. Lastly, an illustration using data from a study of smoking behaviour traits is presented.