This work investigated the use of bioinformatics to predict emulsifying peptides embedded in patatin proteins from potato (Solanum tuberosum). Six peptides (23–29 amino acids) with potentially different predominant structure at the oil/water interface (e.g. α-helix, β-strand or unordered) were identified within patatin sequences. The interfacial tension between peptides solutions and fish oil as well as the physical and oxidative stability of 5 wt% fish oil-in-water emulsions (pH 7) stabilized with synthetic predicted peptides were evaluated. The peptides predicted to have lower amphiphilic score (α1 and α2) led to emulsions with creaming after production and with low oxidative stability. On the other hand, a half hydrophobic and half hydrophilic peptide (γ1), which was predicted to have the highest amphiphilic score, showed a superior ability to reduce interfacial tension (even when compared to casein). γ1-Stabilized emulsion was physically stable during storage (48 h at 50 °C) and presented the lowest droplet size (D_4,3 = 0.518 ± 0.011 μm). Electron spin resonance (ESR) and Oxygraph results indicated that the type of synthetic peptide used also affected the oxidative stability of fish oil-in-water emulsions differently. Therefore, this study shows the potential of using bioinformatics to predict emulsifying peptides, reducing time and cost of extensive screening hydrolysis processes.