Fertilizer use in Indian agriculture is skewed towards nitrogen (N) and phosphorus (P), and farmers generally neglect potassium (K) input ignoring K requirement of crops. We, therefore, studied the long-term effect of imbalanced fertilization (i.e. without K) on K supplying capacity of a kaolinitic red soil (Typic Haplustalf) after 42 years of intensive cultivation. Soil samples (0–15 and 15–30 cm) were collected after the completion of 42 cropping cycles from an on-going long-term fertilizer experiment located at Birsa Agricultural University, Ranchi, India, from six treatments, viz. control (unfertilized), fertilizer N, NP, NPK, 150% NPK (1.5 times of NPK) and NPK + FYM (NPK + farmyard manure), along with adjacent uncultivated land. Soil K pools, clay minerals, quantity-intensity relationship and K release kinetics along with grain yield and K uptake of soybean and wheat were studied. Among the treatments, NP treated soils showed the lowest amounts of labile (K L ) and non-exchangeable K (K nex ). Equilibrium activity ratio (AR e K ) and non-specifically held K (−ΔK 0 ) were also lowest under NP treatment. The gap between equilibrium exchangeable K (EK 0 ) and minimum exchangeable K (EK min ) was minimum under NP. Cropping and fertilization-induced transformation in clay minerals was noticed wherein continuous neglect of K fertilizer led to a decrease in illite content with a concomitant increase in interstratified minerals. On the other hand, noticeably lower alteration in clay minerals and comparatively less depletion in overall K-supplying capacity of soil were recorded under K addition as compared to imbalanced fertilization (i.e. without K). Soybean showed a significant positive response to K application. The study thus underlined the necessity of adequate K input for the sustenance of soil health, inherent K-supplying capacity and crop productivity. © 2019 Elsevier B.V.