During the Bølling–Allerød warm period of the last deglaciation, about 14 kyr ago, there was a strong and pervasive spike in primary productivity in the North Pacific Ocean. It has been suggested that this productivity event was caused by an influx of the micronutrient iron from surrounding continental shelves as they were flooded by sea-level rise. Here we test this hypothesis by comparing numerous proxies of productivity with iron flux and provenance measured from a core from the subarctic Pacific Ocean. We find no evidence for an abrupt deglacial pulse of iron from any source at the time of peak productivity. Instead, we argue that the deglacial productivity peak was caused by two stepwise events. First, deep convection during early deglaciation increased nutrient supply to the surface but also increased the depth of the mixed layer, which pushed surface production deeper in the water column and induced light limitation. A subsequent input of meltwater from northern American ice sheets then stratified the water column, which relieved light limitation while leaving the surface waters enriched in nutrients. We conclude that iron plays, at most, a secondary role in controlling productivity during the glacial and deglacial periods in the subarctic Pacific Ocean.