The purpose of this paper is to estimate popularity functions for the Portuguese Prime Minister, Government, Parliament and President using the ordinary least squares (OLS) and seemingly unrelated regressions (SUR) methods. The results indicate that: (1) popularity polls for the Prime Minister and Government are better explained by economic conditions than are similar polls for the Parliament and President; (2) unemployment is a significant variable determining popularity while inflation is not; (3) honeymoon effects are significant; (4) popularity deteriorates over consecutive terms.