We used the Park Grass Experiment, begun in 1856, to test alternative hypotheses about the relationship between genetic diversity and plant species diversity. The niche variation hypothesis predicts that populations with few interspecific competitors and hence broader niches are expected to contain greater genetic diversity. The coexistence hypothesis predicts that genetic diversity within species favours coexistence among species and therefore species and genetic diversity should be positively correlated. Amplified Fragment Length Polymorphism (AFLP) markers were used to measure the genetic diversity of populations of Anthoxanthum odoratum growing in 10 plots of differing species richness that lie along resource and soil pH gradients. Genetic diversity in A. odoratum was positively correlated with the number of resources added to a plot, but not correlated with species richness. However, separate analyses have shown a negative correlation between resource addition and species richness at Park Grass and elsewhere, so genetic and species diversity appear to respond in opposite directions.