The rice-maize cropping system (RMS) is an emerging option for diversification of the prevalent rice-wheat system (RWS) in South Asia. Studies underlined the significance of adequate potassium (K) nutrition for sustainable intensification of the RMS. Although studies on combined effects of tillage, residue retention, and nutrient management on crop yields and nutrient use efficiency for the RWS are plenty, such studies are scarce for the RMS. We, therefore, conducted a 5-year field study on sandy loam (Typic Ustochrept) soil at Modipuram, India to evaluate the effects of tillage and crop establishment, residue management and K fertilization on crop productivity, K use efficiencies, changes in soil K status, and apparent K balance under irrigated RMS in north-western India. Three combinations of tillage and crop establishment methods i.e., puddled transplanted rice (TPR) followed by conventional-till maize (TPR/CTM); conventional-till dry direct-seeded rice (CTDSR) followed by CTM (CTDSR/CTM); and zero-till DSR followed by zero-till maize (ZTDSR/ZTM) were considered as main plots. Sub-plot treatments consisted of two residue management options i.e., removal of residues (-R) of both crops, and partial residue retention (5 Mg ha−1) for rice and maize, either at soil surface in zero till (ZT) plots or incorporated into the soil in conventionally-till (CT) plots (+R). The sub-sub plot treatments were no-K application (-K) and 62 kg K ha−1 application (+K) to both rice and maize. Data summed-up for 5 years revealed that rice grain yield and K uptake under TPR were significantly higher (p < 0.05) compared with CT/ZT DSR, irrespective of residue and K management options. The maize yield and K uptake were highest under the ZTDSR/ZTM treatment. Both rice and maize yields were significantly (p < 0.05) higher with 62 kg ha−1 K application under −R plots, but such effects were not significant under + R plots. Similarly the effect of residue management on yield and K uptake of rice and maize was significant only in the absence of fertilizer K application. Recycling of rice residue in maize had more pronounced effect on K uptake compared with that of maize residue in rice. Agronomic and recovery efficiencies of applied K were lower under +R plots, and agronomic efficiency was further decreased when both rice and maize were grown under ZT conditions (ZTDSR/ZTM). Soil exchangeable K content at 0–0.05 m and 0.05–0.15 m depths was significantly higher (p < 0.05) under ZTDSR/ZTM compared with that under CTDSR/CTM and TPR/CTM treatments, and the values were greater under +R + K compared with −R−K treatments. Soil exchangeable K under −K−R plots dropped to 102 and 90 mg kg−1 at 0–0.05 m and 0.05–0.15 m depth, respectively vis–a-vis the initial K content (122 and 114 mg kg−1, respectively). The apparent K balance, computed as K addition less K off-take by rice and maize during the study period of 5 years, was invariably negative in the absence of K fertilizer i.e., −307 kg ha−1 in TPR/CTM (+R−K) to −1483 kg ha−1 in ZTDSR/ZTM (−R−K) treatments. On contrary, the apparent K balance was positive in +R + K plots under different tillage and crop establishment methods. Positive correlation between soil exchangeable K and K input indicated that soil K mining in the RMS could be mitigated to a great extent with adequate supply of K through residue retention and K fertilization.