Combining information on the vertical distribution of nutrients and remote sensing can potentially improve estimates of ocean primary production (PP). Here, we employ in situ observations of chlorophyll a and nitrate from biogeochemical Argo floats deployed in the North Atlantic together with remote sensing to estimate PP and compare these results to estimates based on model approaches not including vertically resolved nutrient distributions. Analysis of the float data shows chlorophyll a distribution relates closely to both nutricline depth and latitude, and these relationships can be explained by nutrient and light availability. PP estimates based on satellite and Argo-observations also relate to both latitude and nutrient distributions. An analysis of these float-based PP estimates shows that large-scale patterns of total water column PP and associated variability are consistent with expected photosynthetic responses to different combinations of light and nutrient availability. When PP-estimates based solely on surface observations were plotted against light and nutrient fields, significant structural differences emerged compared with estimates that included subsurface observations, in particular in oligotrophic areas and areas with a shallow nutricline. The combination of in situ water column observations with remote sensing potentially opens a new phase in the estimation of ocean primary production.