Objective: To propose a hypothesis that the long duration of effect of intramuscular (i.m.) vitamin K-1 in preventing late onset haemorrhagic disease results from a depot effect after i.m. injection. Methodology: Review of scientific literature relating to the pharmacology of vitamin K-1 and the aetiology of late onset haemorrhagic disease. Results: A single i.m. dose of vitamin K-1 is effective for at least 2 months, whereas the duration of effect of a single oral dose is about 3-4 weeks. The known pharmacological properties of vitamin K-1 are seemingly at variance with the long duration of effect of an i.m. dose. Menaquinones (vitamins K-2) are absent in the newborn liver, but gradually accumulate after birth. This, together with the low concentrations of vitamin K-1 in human breast milk, may explain the peak frequency of late onset haemorrhagic disease at 4-8 weeks. We hypothesize that after i.m. injection, vitamin K-1 acts as a depot preparation by forming a viscous mass in muscle tissue which is slowly absorbed over many weeks. This hypothesis is supported by reports indicating significantly higher plasma vitamin K-1 levels several weeks after i.m., as compared to oral vitamin K-1. Conclusions: The prolonged efficacy of i.m. vitamin K-1, compared to oral preparations may be due to a depot effect New oral preparations of vitamin K-1, despite greatly improved bioavailability, may have a shorter duration of effect than i.m. vitamin K-1, and therefore be less effective for long-term prophylaxis.