The fraction of total plasma vitamin B(12) bound to transcobalamin (holoTC/B12 ratio) may reflect tissue levels of the vitamin, but its clinical relevance is unclear. We assessed associations between cognitive function and total B12, holoTC, and holoTC/B12 ratio in a cohort of elderly Latinos (n = 1089, age 60-101 years). We assessed cognitive function using the Modified Mini-Mental State Examination (3MSE) and a delayed recall test; we diagnosed clinical cognitive impairment by neuropsychological and clinical exam with expert adjudication; and we assessed depressive symptoms using the Center for Epidemiological Studies Depression Scale (CES-D). We measured total B12 and holoTC using radioassays. HoloTC/B12 ratio was directly associated with 3MSE score (P = 0.026) but not delayed recall score. Interactions between holoTC/B12 and CES-D score were observed for 3MSE (P = 0.026) and delayed recall scores (P = 0.013) such that associations between the ratio and cognitive function scores were confined to individuals with CES-D >/=16. For individuals with CES-D > or = 16, the odds ratio for clinical cognitive impairment for the lowest holoTC/B12 tertile was 3.6 (95% CI 1.2-11.2) compared with the highest tertile (P = 0.03). We observed no associations between cognitive function and total B12 or holoTC alone, except between holoTC and 3MSE score (P = 0.021), and no interactions between holoTC or total B12 and CES-D score on cognitive function. HoloTC/B12 ratio is associated with cognitive function in elderly Latinos with depressive symptoms and may better reflect the adequacy of B12 for nervous system function than either holoTC or total B12 alone.