We investigated the influence of the direction of preceding saccadic trials on the latency of current prosaccades and antisaccades, in healthy subjects and patients with schizophrenia. When prosaccades and antisaccades were performed in separate, single-task blocks, we found that only prosaccades were delayed if the saccade in the prior trial was in the same direction, consistent with the expected directional effect from an 'inhibition of return'-like alternation advantage. However, both types of saccades were executed more quickly when the saccade in the penultimate trial was in the same direction, consistent with previous demonstrations of directional plasticity in monkeys. In blocks of randomly mixed prosaccades and antisaccades, the directional effects in healthy subjects were greatest when a prosaccade was preceded by an antisaccade, consistent with a summation of effects of alternation advantage (from the prior stimulus) and directional plasticity (from the prior saccade). Schizophrenic patients showed an additional phenomenon, a directionally specific inhibition of upcoming saccades by preceding antisaccades. These results suggest that saccades in humans are modulated by inter-trial effects attributable to both an 'inhibition of return'-like alternation advantage and directional plasticity.