Nutrient cycling is an important part of integrated nutrient management. The litterbag method was used in field experiments to determine potassium (K) release patterns from red clover (Trifolium pratense) green manure (GM), field pea (Pisum sativum), canola (Brassica rapa) and monoculture wheat (Triticum aestivum) residues under conventional and zero tillage from 1998 to 1999 and from 1999 to 2000. Potassium contained in crop residues ranged from 25 kg ha-1 in wheat to 121 kg ha-1 in pea residues, both under zero tillage. The percentage of K released over a 52-wk period ranged from 65% of pea K under zero tillage to 99% of clover K under conventional tillage, and the amounts released were 20–32 kg ha-1 from wheat, 31–52 kg ha-1 from canola, 28–79 kg ha-1 from pea, and 31–118 kg ha-1 from legume GM residues. In both trial periods, K from wheat residues was released at a faster rate under conventional tillage than under zero tillage during the first 10 wk of residue decomposition. In contrast, K from pea and canola residues was released more quickly under zero tillage than under conventional tillage. The effect of tillage on K release from GM residues was similar to that on pea and canola residues in 1998–1999, but similar to that on wheat residues in 1999–2000. Correlations between K release and residue quality were inconsistent, presumably because K is not a structural component of plant tissue, and therefore its release is probably related more to leaching than to residue decomposition. These results show that crop residues recycle substantial amounts of K for use by subsequent crops in rotation. Key words: Conservation tillage, crop residue quality, crop rotation, organic soil amendments