Right-wing authoritarianism (RWA) is typically correlated with measures of religiosity. Though there is much cross-sectional evidence for this relationship, few studies have examined the reason for the correlation. Does one variable induce change in the other over time in single individuals, or do they correlate because the same kind of people (as defined by potential third variables) select into both authoritarianism and religiosity? We examine this question with data from a six-wave panel study of Polish adults, using the random-intercept cross-lagged panel model (RI-CLPM) to disentangle within-person relationships from between-person relationships. Though cross-section regressions reveal positive relationships between RWA and two indices of religiosity (attendance and identification as religious), analyses using the RI-CLPM suggest that these relationships are due to between-person differences rather than within-person effects of RWA on religiosity variables and vice versa.