Results are presented from a 30‐year‐old Scots pine provenance trial series covering all Sweden. Variation in total yield is studied and the effect of transfer on this trait. The results show that significant differences exist between provenances in almost all trials. The variance component for provenance effects is about 60% of the total variation. Repeatabilities are found to be high, indicating reliability of data. Based on a response surface model and a contour plotting procedure, transfer functions for southern and northern Sweden respectively, are built. For southern Sweden, a transfer of ±1° combined with an altitudinal downward transfer of 100–200 meters, is proposed. In northern Sweden, no optimum transfer range is found. The function, however, shows that a southward and upward transfer, in combination, gives remarkable gains in total yield.