Mechanisms underlying habitat use in most species are poorly understood. Therefore, we integrated behavioral and forest cover type data to test hypotheses underlying the choice of two key cover types used by male Ruffed Grouse (Bonasa umbellus) in northern Minnesota, USA. Forests of aspen (Populus spp.) are the presumptive highest quality cover type for Ruffed Grouse because this bird species achieves its highest breeding densities in this cover type. Yet, males select cover types of conifer even when nearby cover types of aspen remain vacant. We examined this conundrum – selection of inferior cover types when presumed better types are available – by randomly selecting 23 male Ruffed Grouse from among a contiguous population of territorial males and used automated video systems to monitor their breeding display behavior. We predicted that if conifer cover types were indeed inferior habitat for Ruffed Grouse, males that established territories in these cover types would either drum (auditory display) more frequently to attract mates or have fewer conspecific interactions, detected by observing visual display rates, than males which established territories in aspen cover types. We used drumming rates and visual display rates as two separate response variables, and used generalized linear models to evaluate each of these response variables as a function of several predictors in a priori models including cover type, male density on the study area, distance to the nearest neighbor, cover type of the nearest neighbor, and distance to the nearest neighbor in aspen cover. Information-theoretic model selection was used to rank these models with Akaike's Information Criterion adjusted for small sample sizes. The best predictor (top model) of drumming rates among our model set, was cover type, but our null model was competitive, which suggested uncertainty in our result. We were unable to explain visual display rates given our model set and predictor variables, because our intercept only [null] model was the top model, which suggested visual display behavior was not correlated with either cover type or the other factors we included in models. We conclude the reason grouse select display sites likely has more to do with suitable protective cover (i.e., predator avoidance) than with the general classification (aspen vs. conifer) of the cover type. Thus, the link between relative fitness and habitat selection of specific cover types is more likely to be revealed by studying the habitat requirements and survival probabilities of females visiting males which occupy these two prominent breeding-display covers.