Plant metabolic responses to herbivory are generally elucidated using experiments that compare the metabolite concentrations in attacked plants to un-attacked plants. In these experiments, herbivore density is often arbitrary, and may only elicit one possible set of responses by plants. Furthermore, some metabolites have dose-dependent responses, while others may be induced in an “all-or-nothing” manner. This results in the potential for metabolite blends to change as herbivore density increases and may have important ecological and experimental implications. In nature, plants experience varying degrees of herbivory which influences plant secondary metabolite blends. For example, the ratio of volatile compounds in blends can carry information for parasitoids, and may change tritrophic interactions as herbivore density increases. Experimentally exposing plants to only one density of herbivores may result in an incomplete understanding of downstream effects. Additionally, the quality of many crop plants, such as tea (Camellia sinensis), depends greatly on metabolite blends that are impacted by herbivory. For example, leafhopper damage has been reported to have a positive influence on tea quality through induced production of plant volatiles. We manipulated leafhopper density on tea plants and measured resulting leaf damage as a continuous variable. We sampled volatile and non-volatile metabolites from these damaged tea plants and found metabolite-specific, non-linear associations between leafhopper damage and secondary metabolite concentrations. Furthermore, we found that tea plant genotype influences the effects of leafhopper herbivory on secondary metabolite blends, leading to the potential for different optimal leafhopper densities for improving tea quality.