In order to have expert conclusions on handwriting examinations accepted as evidence in a court of law, forensic handwriting examinations must rely on recognized methodological principles. This paper provides an orientation for expert opinions on handwriting that meets the requirements of scientific method by utilizing hypothesis-guided procedures of inferential statistics and Bayesian logic for obtaining probability statements. The shared wisdom within the scientific community recognizes that expert opinions on handwriting should not be made as categorical statements implying absolute certainty, but rather they should be made as statements of probability. Terms and definitions relating to probability are enumerated, including terms related to the verbal expressions of probability, aspects of measurability, types of probability, logical probability, and probability distribution, among others. Logical models are reviewed that can be used to directly or indirectly derive a probability distribution, such as the Bayesian model and methods of examination that produce a wide range of physical and comparative graphic results. Also considered in this section are the characteristics of hypotheses, types of probabilities, and basic approaches to the derivation and assessment of prior probabilities. In order to minimize the risk associated with verbal misunderstandings, standardization of the probability scale is discussed and includes probability distributions in graduated pairs of verbal, numerically defined expressions of value that each demonstrably adds up to 1. Two case studies are presented, the first illustrates the advantages of relying on Bayes’ approach, while the second uses the case of an abbreviated signature to illustrate the logic of examination and conclusion. The probability model presented in this paper provides the method for arriving at the quantifiable, scientific conclusions demanded of any scientifically-based expert in a court of law. While the discussion here is limited to forensic handwriting analysis, the authors posit that this probability model can be applied to other comparable forensic disciplines. Tables, figures, references