There are subtle signs and symptoms that can function as strong, significant and evidence-based indicators to improve upon diagnosis of mental illness as a primary clinical issue. With the well-being of patients at stake and services experiencing mounting pressure on their time and resources, the accuracy of such a diagnosis is critical. The data used in this study was obtained from the Hampshire Police of individuals arrested under Section 136 of the Mental Health Act 1983 (Amended 2007). Documented observations by the arresting officers were transferred onto a list of signs and symptoms, adapted for clinicians and non-clinicians. Answers to all 23 indicator questions were assigned a binary coding and became the input data. A logistic regression model (LRM), sensitivity and specificity analysis, analysis of maximum likelihood estimates, and confusion matrix, were applied. Three significant evidence-based indicators for primary mental illness diagnosis: “fixed stare/looking (possibly) confused” (perplexed), “calm/seemingly calm”, and “poor attention and concentration, not able to follow instructions easily” (distracted). There was suggestion that when all three identified indicators are absent within the same assessment period, primary mental illness should not be diagnosed. A correlation was observed between the presence of one, two, or all three indicators and the increased chance of patients suffering from moderate-to-severe primary mental illness. The indicators would very likely be particularly helpful in cases where signs and symptoms of primary mental illness are less obvious and made more complex by actual underlying problems of drugs/alcohol, difficult personality traits, and/or criminal behaviour.