Logistic regression modeling is a commonly used method in classification problems. Bayesian methods are becoming easier to implement and are now incorporated into many standard statistical software packages. However, analysts often do not take advantage of the main strength of Bayesian methods and that is their ability to incorporate prior information. In 1996, Bedrick, Christensen, and Johnson introduced a method for soliciting informative priors for logistic regression and other general linear models. In this talk we will discuss a little bit of history and the basics of Bayesian statistics for classification. Then we will introduce the logistic regression model, discuss some problems that can arise in model fitting and learn how some of those problems can be overcome using Bayesian Methods with informative or minimally informative priors. A focus of the talk will be on how to ask the right questions in order to turn knowledge into informative priors that can be utilized to improve classification performance.