Three experiments in which errors of subject-verb agreement were elicited assessed the effects of syntactic function and part of speech of the constituent appearing immediately before the verb. Bock and Miller (Cognitive Psychology, 1991) have shown that constituents modifying the subject exert an "attraction effect", an increased rate of agreement errors when that constituent has a grammatical number different from that of the subject head noun. Experiment 1a, conducted in Dutch, showed that such an attraction effect is not restricted to sentences in which the number mismatching information is embedded within the subject: Direct object noun phrases exert an attraction effect as well, although a smaller one than subject modifier noun phrases. Experiment 1b replicated this effect with new materials, excluding a possible confound with plausibility of the sentences. Experiment 2 showed that direct-object pronouns exert an attraction effect about as strong as observed with nouns, unless the pronoun is explicitly case-marked. In such circumstances no attraction effect obtains. These results are interpreted within the hypothesis that the number of phrasal nodes intervening between "attractor" and subject head noun determines the strength of attraction effects.