Very little is known about the outcomes of poststroke delirium in relation to its symptom spectrum. We investigated the 3-months cognitive and functional outcomes of attenuated (ADS) and full delirium syndromes in Nigerian survivors of first ever stroke. A prospective observational study with repeated assessments conducted in the first week of stroke using the confusion assessment method. Full delirium was diagnosed according to criteria in the fifth edition of the diagnostic and statistical manual of mental disorders (DSM-V). ADS was characterised in survivors who were free of full, but had ≥two core features of, delirium. Baseline and follow-up assessments were conducted using the Mini-Mental state examination (MMSE), 10-words list learning and delayed recall test, Animal naming test and Barthel index. Among 150 participants, ADS was present in 32 (21.3%), full delirium in 29 (19.3%). In linear regression analyses adjusting for age, economic status and systemic hypertension, ADS (Mean difference (MD) = -3.8, 95% C.I = -7.0, -0.7) and full delirium (MD = -5.6, 95% C.I = -9.0, -2.1) independently predicted poorer global cognitive functioning at follow-up. Significant declines in memory (MD = -1.9, 95% C.I = -2.8, 0.9), executive (MD = -2.2, 95% C.I = -4.1, -0.3) and physical functioning (MD = -2.8, 95% C.I = -5.5, -0.2), as well as a 4-fold increase in the independent odds (O.R) for dementia (O.R = 4.1, 95% C.I = 1.0,16.1) were also recorded in full, but not attenuated, delirium. Attenuated and full delirium are associated with graded risk of poststroke cognitive decline. Reconsideration of poststroke delirium as a spectrum, rather than threshold-based categorical diagnosis will improve detection and prioritization of stroke survivors at increased risk of cognitive decline.