Abstract Aim Generative Adversarial Networks (GANs), based on zero-sum game theory, involve two neural networks—the generator and discriminator—to create realistic synthetic data. They're becoming increasingly significant in medicine and plastic surgery. This review assesses GANs' impact in Plastic Surgery, offering a framework for their application and evaluation in various subspecialties, highlighting their potential in research and patient care advancements. Method Following PRISMA guidelines, a systematic review was performed of all applications of GANs within Plastic Surgery and its subspecialties published from 2014 to 2022. Three independent reviewers screened studies using COVIDENCE review management software, from databases including PubMed, Embase, PsychInfo, Scopus and Google Scholar. Results A total of 70 studies were captured by the search, of which seven studies were included in the final analysis. The Kappa score for interrater reliability was 0.97. The most common subspecialty represented was craniofacial (n=4). Proposed uses of GANs ranged from facial recognition, burn estimation, scar prediction and post-breast cancer reconstruction anomaly scoring. All proposed GANs were conditional, with four of the studies using primarily collected training datasets, and the remainder using public datasets. Average dataset training size was 54,593 (range: 300,575-6). All of the studies had different reporting structures, and only four studies had a form of cross-evaluation qualitatively (n=1) or quantitatively (n=3) Conclusions GANs can transform research and patient care. Standardised reporting and diverse dataset transparency are essential. Despite their creation needing computational expertise, it's vital for plastic surgeons to understand GAN development to fully leverage their potential in plastic surgery and related fields.