Abstract Studies have reported beneficial effects of cholinergic enhancers, e.g., rivastigmine, on memory in schizophrenia but others have not. Possibly, these discrepancies are related to the lack of specificity of the tests used. This study investigated the effect of rivastigmine on memory in schizophrenia using event-related potentials (ERPs). Eighteen patients treated with atypical antipsychotic received rivastigmine adjuvant therapy in a randomized, crossover design. They were assessed at baseline (T1) and on two subsequent occasions (T2 and T3), where one half of the subjects were taken rivastigmine and the other half not. ERPs were recorded during a recognition memory task on each session. Behavioral and ERP data were analyzed using mixed ANOVA models first at T1 to detect potential group differences and for the trial (T1–T2) to determine the influence of rivastigmine, i.e., session×group interactions. The results showed no group difference at T1 except a trend for one group to be less efficient than the other on RT measures. When controlling for this difference the results on the trial data showed a trend for a benefit of rivastigmine on the RT memory effect. ERP analysis revealed that rivastigmine affects the amplitudes of two components elicited within 150–300 ms over posterior (reduced N2b) and frontal sites (enhanced P2a). It also enhances the magnitude of the memory (old/new) effect on two later components over posterior (N400) and frontal sites (F-N400). These results suggest that rivastigmine improves selective attention by enhancing interference inhibition processes (P2a) and lowering the reactivity to incoming stimulus (N2b). It also improves the integration of information with knowledge (N400) and with its context (F-N400). Generally, this study showed that the beneficial effect of rivastigmine on memory is not unitary but rather comes from its action at different time points within information processing cascade.