1. During the course of a nutritional survey on an Antarctic base, leucocyte ascorbic acid estimations (Denson & Bowers, 1961) were found to be significantly affected (P< 0.0001) by the leucocyte count measured during analysis.2. The effect was significantly less (P< 0.05) when leucocyte levels were at or near saturation than when subsaturated.3. This suggests that leucocytes share available ascorbic acid, especially in subsaturated subjects, and that assessment of nutritional status from the leucocyte ascorbic acid level should take account of the leucocyte count and the plasma ascorbic acid titre.