The Jan Mayen shelf has recently been identified as a spawning area for cod (Gadus morhua). Genetic analyses and otolith types of adult cod caught in the Jan Mayen area show that they are of mixed origin, from both Iceland and the Barents Sea. However, they are genetically different from both stocks indicating the possibility of a cod stock around Jan Mayen. Motivated by this, we investigate the potential for eggs and larvae of cod to stay behind at Jan Mayen after spawning, which would be a requirement for having a self‐sustained stock. This is done using a numerical model for larval drift and growth, similarly to what has been done for other cod stocks. Using the modelled drift, we compare the proportion of larvae retained in different years to the strength of year classes as indicated by fisheries data. We find that even if most individuals are transported away from Jan Mayen, there are still some years where a significant proportion (more than 20%) of the larvae remains at the time of settlement in October–November. However, the year‐to‐year variability is large, with almost no individuals remaining in certain years. Our results suggest that aggregated over years, the amount of larvae that reach the age of bottom settlement on the Jan Mayen shelf area could be sufficient to give a potential for a self‐sustained cod population. In addition, we find that the main drift direction away from Jan Mayen is southwest towards East Greenland.