A considerable challenge in applied agricultural use of reflection-based spectroscopy is that most analytical approaches are quite sensitive to radiometric noise and/or low radiometric repeatability. In this study, hyperspectral imaging data were acquired from individual maize leaves and the main objective was to evaluate a classification system for detection of drought stress levels and spider mite infestation levels across maize hybrids and vertical position of maize leaves. A second objective was to estimate biomass and biofuel potential (heating value) of growing maize plants. Stepwise discriminant analysis was used to identify the five spectral bands (440, 462, 652, 706, and 784 nm) that contributed most to the classification of three levels of drought stress (moderate, subtle, and none) across hybrids, leaf position, and spider mite infestation. Regarding the five selected spectral bands, average reflectance values and standard variogram parameters ("nugget", "sill", and "range" derived from variogram analysis) were examined as indicators of spider mite and/or drought stress. There was consistent significant effect of drought stress on average reflectance values, while only one spectral band responded significantly to spider mite infestations. Different variogram parameters provided reliable indications of spider mite infestation and drought stress. Based on independent validation, variogram parameters could be used to accurately predict spider mite density but were less effective as indicators of drought stress. In addition, variogram parameters were used as explanatory variables to predict biomass and biofuel potential of individual maize plants. The potential of using variogram analysis as part of hyperspectral imaging analysis is discussed.