This study extended Brown and Levinson's model of politeness in order to explain more complicated forms of interaction and proposed three different types of facework: solidarity, approbation, and tact. The effects of three social factors—relational intimacy, power difference, and the right to perform a given act in a given situation—on the three types of facework were examined. Relational intimacy was the strongest positive predictor of facework. Its effect was consistent across different types of facework and across different levels of power and right. The right to perform a certain act and speaker's power decreased facework, but these effects were mediated by relational intimacy, achieving significance mainly in distant relationships. The study also examined the relationships among different types of facework. Respondents used multiple types of facework when multiple face wants were threatened, and the use of one type of facework did not decrease the use of other types.