Potato yields are influenced by the spatial variability of soil texture and organic matter content. Variable rate application (VRA) of fertilizers by management zone (MZ) or on a continuous basis (VRAC) can help manage this soil variability. Experiments had been conducted in eastern Canada under rain-fed potato production to evaluate different VRA strategies of N, P, K and papermill residues (PR). The first experiment evaluated VRA of P and K in a two-ha field by MZ based on P and K soil test maps. The second experiment evaluated VRA of PR in a 24-ha field by MZ based on a soil organic matter map. The two last experiments evaluated VRA of N using MZ delineated with soil electrical conductivity map and using VRAC of N based on potato N status as detected with the Hydro N Sensor. The VRA of P and K was unsuccessful and showed that yield variation was mostly explained by field pedodiversity. The absence of P and K soil specific recommendations were also a limitation. The VRA of PR significantly improved potato yields mainly due to the increase of soil water holding capacity. The two MZ delineated with soil EC were significantly different in terms of yield potential, and in some cases crop response to N application strategy differed between MZ. The Hydro N Sensor was useful for mapping the spatial variability of crop N status; however, the success of VRAC of N was limited by the uncertainty in N recommendation associated with the crop N status. To date, no study in eastern Canada has attempted to combine these two VRA approaches. Using MZ to optimize soil management in combination with VRAC during the growing season based crop status may be the most sustainable way to implement precision agriculture in potato production.