A rice-soybean rotation has been the dominant cropping system in eastern Arkansas, but emerging water and disease issues may promote in places the cropping of continuous rice. In a five-year field experiment, grain yields were 19% less for continuous rice than for rice following soybean. To identify the cause of the yield gap, ¹⁵N-labeled fertilizer was applied in 2002 and 2004 to microplots imbedded in this field. All N fertilizer was added as the standard single application early in the growing season. Results attributed the yield gap mostly to decreased uptake of unlabeled N by continuous rice compared to rice following soybean. The unlabeled N had presumably been mineralized from soil organic matter. The difference between rotations in uptake of ¹⁴N became most pronounced in mid- to late-season, when it reached 25 kg N ha^-1. Crop uptake of ¹⁵N-labeled fertilizer N was about 10 kg N ha^-1 less for continuous rice than for rice following soybean. Lignin-derived phenols in soil decomposed more slowly in-season for continuous rice than for rice following soybean. This association of inhibited soil N mineralization with an enrichment of soil phenols is consistent with the hypothesis that under the relatively anaerobic conditions of the continuous rice rotation, incompletely decomposed lignin compounds from crop residues chemically bind soil N into less available forms. A similar finding was reported from long-term rice studies in the Philippines. Mitigation options to be evaluated include multiple applications of N fertilizer and aerobic decomposition of crop residues to promote decomposition of phenolic compounds.