All topics in ecology and conservation play out in space and time, and scale describes the spatio-temporal domain of a pattern or process. By understanding and quantifying scale, it can profoundly influence our understanding of ecological patterns and processes, and it can alter conservation decisions. In addition, many patterns and processes occur at fundamentally different scales in space and time. Our goals are to introduce concepts for understanding spatial scale, provide examples of why scale is important, and illustrate some simple approaches to multi-scale and multi-level modeling with an example of reptile distribution (i.e., the five-lined skink, Plestiodon inexpectatus) in the Southeastern United States. Scale has two primary components, grain and extent, and it can be applied to describe ecological phenomena, ecological sampling, or data analysis. These issues have led to some confusion in terms of the concept of scale and its application. Spatial scale has been shown to be important for ecological and conservation problems because ecological processes can operate at different scales and patterns can be fundamentally different depending on the scale at which they are quantified. Two general approaches to the problem of scale include multi-scale and multi-level modeling. Multi-scale modeling quantifies environment conditions at multiple scales by altering either the grain or extent of the analysis, and then evaluates which of the considered scales best explains a pattern or process. In multi-level modeling, the focus is on interpreting effects at different levels in an organizational hierarchy. We illustrate the use of buffer-based and kernel-based multi-scale modeling of species–environment relationships. Our example shows how to optimally identify the scale at which species respond to the environment but also illustrates how the quantification of pattern is highly correlated across scales, a common issue in multi-scale modeling. We end by discussing some further advances in understanding and quantifying scale.