One way to assess military perceptual expertise is to present brief stimuli to military experts and novices and to ask them to make a 2-choice expertise-related judgment as quickly and accurately as they can. The differences in the distributions of their reaction times and errors can then reveal differences in their perceptual expertise. The difficulty is that reaction times and errors are often not independent. In some cases, participants make a speed-for-accuracy tradeoff. In other cases, correct responses occur quickly but error responses require a lot of perceptual processing. Either way, an analysis of reaction times or errors by themselves can produce misleading results. An increasingly popular solution is to use reaction time and error distributions together to perform a Drift Diffusion Model (DDM) analysis. This approach yields a profile of several cognitively meaningful components, including an estimate of the speed of processing, the level of the response threshold (indicating how much information the participants needed in order to make a perceptual decision), and the amount non-decision time, which often translates to the amount of time it takes to encode the stimulus. By comparison, traditional methods for assessing reaction time and accuracy (separately) do not differentiate the component cognitive processes of encoding, decision-making, and response execution, respectively. These components can yield surprising results. For example, DDM has shown that the main culprits in aging are increases in response threshold and in non-decision time but not speed of processing. Moreover, speed of processing, but not response threshold or non-decision time, is related to working memory capacity and reasoning ability in adults of all ages. In this paper we will discuss strategy and techniques of DDM analysis. We will then illustrate them by discussing an experiment we performed with military experts and novices using this methodology, with the aim of encouraging other researchers to adapt the approach for their own research issues.