Assisted gene flow (AGF) may help facilitate tree species’ adaptation to future climatic conditions. When applied to resource-producing species such as black spruce (BS, Picea mariana), AGF could contribute to maintaining forest productivity and therefore carbon sequestration. For AGF to succeed, it is important to determine how growth varies among different populations under different climates, especially for species with large distribution ranges. Existing common garden experiments represent a unique opportunity to assess and guide AGF. Using BS common garden experiments established across Canada in the seventies, the objectives of our study were to assess: 1) which black spruce population is best suited for AGF and 2) which climatic factors are the most influential in terms of growth. We achieved this by modelling the growth of black spruce using an artificial neural network trained with dendroecological and genetic data obtained from more than 2600 trees (> 80 provenances) growing in four common gardens. When simulating growth until 2100 using the common gardens as planting sites and climate projections following RCP4.5, populations from the Maritimes are amongst the most productive in all gardens except Alberta. In the latter case, local provenances are the most productive. As a general trend, populations that perform the best are from the same latitude or slightly lower latitudes. Climatic variables having the most importance in these predictions were autumn evapotranspiration and autumn mean air temperature (current and previous year of growth). In conclusion, it seems that for black spruce, AGF could be more interesting for migration from maritime to continental climates rather than from south to north.