Assessing vegetation nutrient limitation is a fundamental part of understanding the dynamics of grasslands and savannas. The ratio of concentrations of N and P in vegetation (N:P) has been proposed as an index of the relative limitation of biomass production by the two nutrients. Although successfully used to predict nutrient limitation for wetlands and considered to work for upland ecosystems, we show that N:P appears to be a poor predictor of nutrient limitation in upland grasslands. At five sites in Kruger National Park in South Africa, across soil and precipitation contrasts, we added N and P in factorial to grass-dominated plots. Although across all sites the N:P of unfertilized vegetation (5.8) would have indicated that production was limited by N, aboveground production was consistently co-limited by N and P.  After N fertilization, aboveground production still responded to P, but N:P had only had increased to 9.9, well below standard thresholds for P limitation. Comparisons among sites showed little pattern between site N:P and relative responses to N and P. When joined with results from other fertilization studies in grasslands, caution should be used in interpreting nutrient concentration ratios in upland grasslands as representing nutrient limitation. Co-limitation between N and P appears to be much more widespread than would be predicted from simple assumptions of nutrient supply and limitation. This will complicate our ability to inform ecosystem stewards as to the effects of fire, herbivory, and nutrient deposition on ecosystem processes until we better understand nutrient cycling in grasslands.