Lignin’s role in litter and soil organic carbon (SOC) decomposition remains contentious. Lignin decomposition was traditionally thought to increase during midstage litter decomposition, when cellulose occlusion by lignin began to limit mass loss. Alternatively, lignin decomposition could be greatest in fresh litter as a consequence of co‐metabolism, and lignin might decompose faster than bulk SOC. To test these competing hypotheses, we incubated 10 forest soils with C4 grass litter (amended with 13C‐labeled or unlabeled lignin) over 2 yr and measured soil respiration and its isotope composition. Early lignin decomposition was greatest in 5 of 10 soils, consistent with the co‐metabolism hypothesis. However, lignin decomposition peaked 6–24 months later in the other five soils, consistent with the substrate‐limitation hypothesis; these soils were highly acidic. Rates of lignin, litter, and SOC decomposition tended to converge over time. Cumulative lignin decomposition was never greater than SOC decomposition; lignin decomposition was significantly lower than SOC decomposition in six soils. Net nitrogen mineralization predicted lignin decomposition ratios relative to litter and SOC. Although the onset of lignin decomposition can indeed be rapid, lignin still presents a likely bottleneck in litter and SOC decomposition, meriting a reconsideration of lignin’s role in modern decomposition paradigms.