Background: Suicidality could be associated with specific combinations of biological, social and psychological factors. Therefore, depressive episodes with suicidal ideation could be different from depressive episodes without suicidal ideation in terms of latent variable structures. Methods: In this study we compared latent variable structures between suicidal and non-suicidal depressed patients using confirmatory factor analysis (CFA), following a measurement invariance test procedure. Patients (N = 919) suffering from major depressive disorder were selected from the Netherlands Study of Depression and Anxiety (NESDA) and split into a group that showed no symptoms of suicidal ideation (non-SI; N = 691) and a suicidal ideation (SI) group that had one or more symptoms of suicidal ideation (N = 228). Depression and anxiety symptoms were measured using the short form of the Mood and Anxiety Symptoms Questionnaire (MASQ-D30). Results: CFA implied a difference in latent variable structures between the non-SI sample (CFI 0.957; RMSEA 0.041) and the SI sample (CFI 0.900; RMSEA 0.056). Subsequent multiple-group CFA showed violations of measurement invariance. The General distress and Anhedonic depression subscales were best indicated by hopelessness and lack of optimism in the SI sample and by dissatisfaction and not feeling lively in the non-SI sample. Overall, the SI sample had higher scores and lower inter-item correlations on the Anhedonic depression items. Limitations: We have included very mild cases of suicidal ideation in our SI sample. Conclusions: On a latent variable level, depression with suicidal ideation differs from depression without suicidal ideation. Results encourage further research into the symptom structure of depression among suicidal patients.