Research on persuasion has shown that inferences based on heuristic or peripheral cues can bias the subsequent processing of persuasive messages. Two studies (total N ¼ 296) examined the additional possibilities that a message argument can serve as a biassing factor and cue-related information can serve as the target of processing bias. It was demonstrated that a message argument can bias (a) the processing of subsequent other message arguments (Study 1) and (b) the processing of subsequent cue information (Study 2). Results are discussed within dual-process models and the recently developed unimodel of persuasion. Copyright # 2007 John Wiley & Sons, Ltd.Significant improvements in our understanding of how people form and change attitudes in response to persuasive communication have been accomplished by dual-process models of persuasion, namely Petty and Cacioppo's 'Elaboration Likelihood Model' (ELM; e.g. Petty & Cacioppo, 1986) and Chaiken and colleagues' 'Heuristic-Systematic Model' (HSM; e.g. Chaiken, Liberman, & Eagly, 1989). Both models hold that under certain conditions the processing of information relevant to the issue under consideration (termed central route and systematic processing, respectively) can be biassed by factors exogenous to the message like source characteristics (e.g. Bohner, Ruder, & Erb, 2002;Chaiken & Maheswaran, 1994), recipients' mood states (Bohner, Chaiken, & Hunyadi 1994; Petty, Schuman, Richman, & Strathman, 1993), consensus among proponents (Darke et al., 1998;Erb, Bohner, Schmälzle, & Rank, 1998), and others (e.g. Chen, Shechter, & Chaiken, 1996;Ziegler, von Schwichow, & Diehl, 2005). Within this framework, the present article explores previously European Journal of Social Psychology Eur. J. Soc. Psychol. 37, 1057-1075(2007 According to both the ELM and the HSM, biassed processing reflects an effect of some peripheral or heuristic factor on the valence of cognitive responses toward message content and subsequent attitude judgement. For example, Chaiken and Maheswaran (1994) observed that cognitive responses toward a message on a telephone answering machine were more positive when the communicator was perceived to be of high credibility (a magazine specialized in scientific product testing) versus low credibility (a promotional pamphlet of a discount department store chain). The effect occurred when (a) recipients were sufficiently able and motivated to process the persuasive message thoroughly and (b) the message itself was neither clearly weak nor strong, but sufficiently ambiguous to allow the biassing factor to contort its content. Biassed cognitive responses to the message, in turn, led to favourable or unfavourable attitude judgements toward the consumer product. The assumed underlying mechanism was that heuristic-based initial pre-judgements, positive (negative) in the case of a highly credible (incredible) source, coloured the valence of cognitive responses that recipients finally based their judgements on (Chaiken & Maheswaran, 1994, p. 461). Biassed processin...