Observations over a period of 39 days of the increasing minimum core temperature of Winter Water trapped within a mature cyclonic eddy in the Antarctic Circumpolar Current encouraged us to assess the mixing processes which might lead to this temperature rise. Using diffusive heat budgets for alternatively diapycnic (∼vertical) and then isopycnic (∼horizontal) mixing combined with the observed rate of warming allowed upper limits for diapycnic and isopycnic diffusivities to be inferred. This gave values of (3.3 ± 0.8) × 10−4m2s−1 and 87 ± 20m2s−1 for the diapycnic and isopycnic diffusivities respectively. These were then in turn applied to the isopycnic temperature distributions of a juvenile eddy observed earlier in the cruise and integrated forward in time. While both forms of diffusion undoubtedly play a role in modifying the temperature, it was the horizontal diffusivity which was better able to reproduce the tighter θS relationship and the horizontal spread of the temperature minimum observed in the mature eddy.