This project aims to make a broad characterization of agreement encoding processes, both during language production and comprehension, by proposing a cross-linguistic research project. We aim to characterize verb agreement encoding with its arguments by means of the number attraction phenomenon. Most psycholinguistic research has only considered subject-verb agreement, mainly because it is the only form of verb agreement that exists in the languages that have been studied (e.g., English, Dutch, French, Italian). However, we will further explore number agreement processes of the verb and its other arguments (e.g., subject, object and indirect object) both in Basque and Spanish. In Basque the verb agrees with the verb with its three arguments. In Spanish, the verb only agrees with the subject. However, it has also been suggested that object-clitic agreement is a kind of object-verb agreement (e.g. Franco, 1993).These issues will be explored by means of behavioural agreement error elicitation tasks in production, and event-related brain potentials (ERPs) in comprehension. Our working hypothesis is that similar general processing mechanisms underlie different agreement relations between the verb and its different arguments. We expect to find signs of different sensitivity to attraction effects between different agreement relations (e.g., larger/shorter attraction effects in production; differences in the amplitude, distribution and/or latencies of the ERP components). This would suggest that the strength of these relations during on-line production and comprehension processes may be different (e.g., subject agreement relations being more robust and less prone to attraction effects than object/dative agreement relations).