Three studies tested the hypothesis that assimilation of impressions to primed constructs is a product of relatively superficial processing and is unlikely to occur when behavioral information about a target person is processed systematically. In Study 1, the impressions of accuracy-motivated Ss did not assimilate to covertly primed trait constructs, although the impressions of unmotivated Ss did. Studies 2 and 3 showed that when Ss become accuracy motivated after exposure to target information, both retrieval of that information and opportunity for effortful processing of it were necessary to eliminate assimilation effects. In addition, accuracy-motivated Ss showed no special attention to primes or awareness of their influence on judgment. Even when one is unaware of the potential biasing influence of primed constructs, they often may not bias impression formation, so long as available target information is processed systematically.