Many species around the world have collapsed, yet only some have recovered. A key question is what happens to populations post collapse. Traditionally, marine fish collapses are linked to overfishing, poor climate, and recruitment. We test whether the effect on biomass change from these drivers remains the same after a collapse. We used a regression model to analyse the effect of harvesting, recruitment, and climate variability on biomass change before and after a collapse across 54 marine fish populations around the world. The most salient result was the change in fishing effect that became weaker after a collapse. The change in sea temperature and recruitment effects were more variable across systems. The strongest changes were in the pelagic habitats. The resultant change in the sensitivity to external drivers indicates that whilst biomass may be rebuilt, the responses to variables known to affect stocks may have changed after a collapse. Our results show that a general model applied to many stocks provides useful insights, but that not all stocks respond similarly to a collapse calling for stock-specific models. Stocks respond to environmental drivers differently after a collapse, so caution is needed when using pre-collapse knowledge to advise on population dynamics and management.