Suicide-preventive training has shown to influence attitudes. This study aimed at investigating what impact other factors than knowledge might have on attitudes towards work with suicidal patients and suicide prevention. In 2007, 500 health-care staff working in a psychiatric clinic in Stockholm received a questionnaire with items concerning work with suicidal patients to which 358 (71.6%) responded. A set of attitude items were tested using structural equation modelling (LISREL). Three models were found to be satisfactory valid and reliable: Job clarity, Job confidence and Attitudes towards prevention. These were then used in regression analyses as dependent variables with predictors such as experience of work with suicidal patients, perceived sufficient training, age and gender. Perceived sufficient training was consistently the most important predictor for all three attitude concepts (p < 0.01, β = 0.559 for Job clarity; p < 0.01, β = 0.53 for Job confidence; p < 0.01, β = 0.191 for Attitudes towards prevention). Age was another significant predictor for Job clarity (p < 0.05, β = 0.134), as was experience of patient suicide for Job confidence (p < 0.05, β = 0.137). It is concluded that providing suicide preventive education is likely to improve attitudes towards the prevention of suicide, clarity and confidence regarding their role in the care for suicidal patients. These improvements may contribute to the prevention of suicide in health care settings.