Marine fish and invertebrates are shifting their regional and global distributions in response to climate change, but it is unclear whether their productivity is being affected as well. Here we tested for timevarying trends in biological productivity parameters across 262 fish stocks of 127 species in 39 large marine ecosystems and high-seas areas (hereafter LMEs). This global meta-analysis revealed widespread changes in the relationship between spawning stock size and the production of juvenile offspring (recruitment), suggesting fundamental biological change in fish stock productivity at early life stages. Across regions, we estimate that average recruitment capacity has declined at a rate approximately equal to 3% of the historical maximum per decade. However, we observed large variability among stocks and regions; for example, highly negative trends in the North Atlantic contrast with more neutral patterns in the North Pacific. The extent of biological change in each LME was significantly related to observed changes in phytoplankton chlorophyll concentration and the intensity of historical overfishing in that ecosystem. We conclude that both environmental changes and chronic overfishing have already affected the productive capacity of many stocks at the recruitment stage of the life cycle. These results provide a baseline for ecosystem-based fisheries management and may help adjust expectations for future food production from the oceans.fisheries | population dynamics | productivity | recruitment | nonstationary processes H uman well-being is closely linked with the productivity of marine fisheries, which provide a significant source of protein for more than half of the world's population (1). However, ongoing environmental and biological changes may impact productivity through a variety of mechanisms, including larger habitat areas for temperate species (2), altered body sizes (3), food availability (4), and increased exposure to oxygen-depleted and acidic waters (5). Recent research has documented marked changes in the distributional patterns of marine species that are consistent with climate forcing (6, 7). However, the net effect of these changes on global fish stock productivity is not clearly understood. In particular, documented environmental changes (4,8,9) and the long-term consequences of overfishing (10, 11) all impose relevant but poorly constrained effects. Here we help address this issue by evaluating the evidence for empirical trends in the relation between the size of the reproductively mature population (or "spawning stock") and the annual production of juvenile offspring ("recruits") using a recently synthesized global database of stock-recruit time series (12). We then test the relation between empirical recruitment trends and regional environmental variables associated with temperature, phytoplankton abundance, and historical overfishing.Recruitment is modeled by relating the size of the spawning stock biomass to the annual production of recruits. The magnitude of annual recruitment is hi...