Phenotypic expression depends on both the underlying genetics and the environment the phenotype is expressed in, i.e., plasticity. Adaptive theory predicts that selection should align with the dimensions of most genetic variation and plasticity because this will increase the evolutionary rate of a population, meaning that a population would reach its fitness optimum faster than if they were misaligned. Alignment with selection is only predicted if there is directional selection, and not under stabilizing selection. In addition, only adaptive plasticity is predicted to align with both selection and genetic variation, with the proportion of the plastic variation consisting of adaptive plasticity determining how well aligned plasticity should be. In the first chapter of this dissertation, I outline the evolutionary consequences of the relationship between selection, genetic variation, and plasticity, as well as what the predictions are for their alignments and how to estimate them. In my second chapter I empirically test the alignment between selection, among- and within-individual variation (used as proxies for genetic variation and plasticity respectively) for three behaviors in a wild population of deer mice (Peromyscus maniculatus). I found that selection, among- and within-individual variation were all misaligned, and that there was very little variation in all three behaviors. This could indicate that the behaviors have already reached their fitness optimum due to previous selection pressure. Consequently, this population might not be able to adapt to environmental change. In my last chapter I investigate the cost of plasticity in response to a predatory cue on reproductive outputs in isogenic lines of the banded cricket (Gryllodes sigillatus). Plasticity is assumed to have associated costs which would affect its alignment with selection and genetic variation. I found no evidence for cost of plasticity in G. sigillatus, and in addition there was no genetic variation in plasticity among the lines. Again, previous selection might drive the population’s mean plasticity to its fitness optimum, reducing the variation and the costs of plasticity, making it harder to detect.