Forty two patients who had unilateral temporal lobe surgery (either temporal lobectomy or amygdalo-hippocampectomy) were evaluated using a selection of cognitive tests before and soon after surgery, to examine whether the amygdalo-hippocampectomy produces less cognitive impairment than the standard en bloc resection. On specific indices of cognitive functioning an amygdalo-hippocampectomy rather than a temporal lobectomy, undertaken on the temporal lobe thought to mediate that particular function, produced less impairment, in terms of change in cognitive function resulting from the operation. An amygdalo-hippocampectomy carried out on the temporal lobe not thought to mediate such skills, however, resulted in less improvement or more deterioration in these functions than a temporal lobectomy, except in the case of delayed prose recall, where a right amygdalo-hippocampectomy led to more improvement than a right temporal lobectomy. Overall there were few scores which distinguished between the different surgical procedures for cognitive outcome.