ObjectiveTo evaluate morbidity during long‐term follow‐up with clinical treatment of affective and schizoaffective disorder subjects followed from hospitalization for first major psychotic episodes.MethodsWe followed adult subjects systematically at regular intervals from hospitalization for first‐lifetime episodes of major affective and schizoaffective disorders with initial psychotic features. We compiled % of days with morbidity types from detailed records and life charts, reviewed earliest antecedent morbidities, compared both with final diagnoses and initial presenting illness types, and evaluated morbidity risk factors with regression modeling.FindingsWith final diagnoses of bipolar‐I (BD‐I, n = 216), schizoaffective (SzAffD, 71), and major depressive (MDD, 42) disorders, 329 subjects were followed for 4.47 [CI: 4.20–4.47] years. Initial episodes were mania (41.6%), mixed states (24.3%), depression (19.5%), or apparent nonaffective psychosis (14.6%). Antecedent morbidity presented 12.7 years before first episodes (ages: SzAffD ≤ BD‐I < MDD). Long‐term % of days ill ranked SzAffD (83.0%), MDD (57.8%), BD‐I (45.0%). Morbidity differed by diagnosis and first‐episode types, and was predicted by first episodes and suggested by antecedent illnesses. Long‐term wellness was greater with BD‐I diagnosis, first episode not mixed or psychotic nonaffective, rapid onset, and being older at first antecedents, but not follow‐up duration.ConclusionsInitially, psychotic BD‐I, SzAffD, or MDD subjects followed for 4.47 years from first hospitalization experienced much illness, especially depressive or dysthymic, despite ongoing clinical treatment. Antecedent symptoms arose years before index first episodes; antecedents and first episode types predicted types and amounts of long‐term morbidity, which ranked: SzAffD > MDD > BD‐I.