Objective: Prior research suggests that approximately two‐thirds of public mass shooters exhibit signs of mental illness. This study analyzed whether that means there are two psychological types of perpetrators (some mentally ill, some mentally healthy), or whether almost all perpetrators are likely to have mental health problems. Method: Using a database of 171 public mass shooters who attacked in the United States from 1966‐2019, we tested for statistically significant differences between perpetrators with and without diagnoses or signs of mental illness. We also closely examined the most lethal perpetrators since 2012, and the most “mentally healthy” perpetrators according to prior coding. Results: Correlates of mental illness were approximately equally common among perpetrators, whether they were believed to be mentally ill or not. Of the variables we examined, data availability provided the best explanation for coding of mental illness, not any trait or life experience. Further evidence suggested that even the most “mentally healthy” perpetrators could be recoded as having signs of mental illness or suicidality, or were clear outliers, or may not qualify as public mass shooters. The most lethal perpetrators exhibited signs of mental illness or suicidal intent (or both) in all cases. Conclusion: When people engage in concerning behaviors that suggest a mass shooting risk, their mental health should be carefully assessed alongside other warning signs. However, it is important to avoid treating people with mental illness like criminals, because social stigma reduces the likelihood that they will ask for, and receive, the psychological help they need.