Stagnating potato (Solanum tuberosum L.) yields in eastern Canada have resulted in loss of competitive advantage in global potato markets. Therefore, there is a need to investigate the potential to increase yield by adopting precision agriculture technology. This study evaluated the efficiency of an apparent soil electrical conductivity (EC_a) sensor to delineate management zones (MZs) in two commercial potato fields in New Brunswick, Canada, using an unsupervised fuzzy k-means clustering algorithm. Georeferenced soil samples from 0 to 15 cm depth were analyzed for physicochemical properties. Tuber yields were recorded using a yield monitor. The two MZs delineated using soil EC_a differed significantly in soil physicochemical properties for both fields; however, tuber yield differed significantly between MZs only in Field 1. The yield difference (7.1 Mg ha^-1) in Field 1 was attributed to a difference in soil moisture (23.5% vs 28.5%) resulting from a difference in clay content (141 vs 189 g kg^-1). The lack of a yield difference between MZs in Field 2 may reflect relatively low within-field spatial variability. The soil EC_a sensor showed promise for use in commercial potato production in New Brunswick, especially in fields with high spatial variability.