Three experiments tested the Martens, Vealey, and Burton (1990) model of competitive anxiety. In experiment one, perceived uncertainty of outcome, one of three hypothesised causes of state anxiety (A-state), was manipulated by assigning 72 golfers to either a low uncertainty group (LU), composed of unequal ability pairs, or a high uncertainty group (HU), consisting of equal ability pairs. In experiment two perceived importance, another hypothesised cause of A-state was manipulated by assigning 52 club golfers to either a low importance (LI) group that competed for three golf balls, or a high importance (HI) group that competed for a new pair of golf shoes. Experiment three tested the hypothesised interaction between perceived uncertainty and importance by manipulating both perceived uncertainty and importance with 100 golfers.