We studied the continuity of intergroup bias across two generations among the Swedishspeaking Finns. Many moderators influence the intensity of intergroup bias, for example, group size, group status, valence condition (i.e., attribution of positive or negative stereotypes to in-and outgroup) and ingroup identification. In multilingual contexts, the ethnolinguistic vitality of a language group can be conceived as a proxy for its status. The socalled 'aggravating hypothesis' holds that inferior group size or group status constitute aggravating conditions which eliminate the so called PNAE ('positive-negative asymmetry effect'). We found, in accordance with this hypothesis, that neither generation of this minority group showed PNAE, i.e., both generations showed ingroup favouritism in the negative (attribution of negative stereotypes) as well as in the positive (attribution of positive stereotypes) stimulus conditions. In addition, intergroup bias was stronger among those with a strong ingroup identity, and weaker among those with a strong national identity, regardless of age, although the adolescents showed less intergroup bias than the adults in the negative stimulus condition. This difference occurred only after controlling for ingroup identity, as the adults identified with their ingroup significantly more than the adolescents. However, the adolescents favoured their ingroup more than the adults in the positive stimulus condition. Low local objective ethnolinguistic vitality (i.e. small proportion of Swedish-speakers in the community) seemed to imply an additional aggravating condition, as it predicted more intergroup bias in both generations and in both valence conditions. In accordance with expectations, however, a strong interpersonal network of language contacts, i.e., only Swedish spoken at home, was related to stronger intergroup bias in both generations and both valence conditions.