One of the most critical modifications affecting the N-terminus of proteins is N-myristoylation. This irreversible modification affects the membrane-binding properties of crucial proteins involved in signal transduction cascades. This cotranslational modification, catalyzed by N-myristoyl transferase, occurs both in lower and higher eukaryotes and is a validated therapeutic target for several pathologies. However, this lipidation proves very difficult to be evidenced in vivo even with state-of-the-art proteomics approaches or bioinformatics tools. A large part of N-myristoylated proteins remains to be discovered and the rules of substrate specificity need to be established in each organism. Because the peptide substrate recognition occurs around the first eight residues, short peptides are used for modeling the reaction in vitro. Here, we provide a novel approach including a dedicated peptide array for high-throughput profiling protein N-myristoylation specificity. We show that myristoylation predictive tools need to be fine-tuned to organisms and that their poor accuracy should be significantly enhanced. This should lead to strongly improved knowledge of the number and function of myristoylated proteins occurring in any proteome.