This study examined suicide assessment validity by comparing methods of measuring current risk associated with past suicidal behaviors. Three independent samples (Ns = 359, 1007, and 713; aged 18–76 years) all included participants covering a broad spectrum of suicidality. Information theory, item response theory, general linear modeling, and linear regression modeling tested seven competing methods/models of assessing past suicidal behaviors in relation to current suicidality. In contrast to contemporary theories, ANOVA results showed suicide plans can indicate higher risk than suicide attempts when intent to die is higher. Contrary to popular practice, evidence demonstrated that defining risk by suicide ideation (yes/no), attempts or serious attempts (yes/no), are false dichotomies, were the least valid models tested, and failed to explain substantial explainable variance in suicidality/risk. A newly proposed model, differentiating behaviors with or without intent to die, was the most efficacious dichotomous method. However, as predicted, continuous variables were superior to dichotomous. The proposed suicidal barometer model (SBM) exhibited robust evidence as the best available model for evaluating suicidal behaviors in all samples (100 % probability), explaining 47–61 % of suicidality variance and provided incremental improvement in risk evaluations. Findings were consistent by sample, sex, age-group, ethnicity, and psychiatric history. This study, and related evidence, demonstrate that there is a clear and present need for updating measures, clinical training and core competencies, for valid assessment and risk formulation.