This chapter adapts competitive balance ratio to measure the effects of game-level outcome uncertainty on soccer attendance. The Hausman model selection test is applied to choose the appropriate specification of censored or uncensored interval regression model. The empirical analysis utilizes data from the South African Premier Soccer League’s 1200 games over five seasons, 2010–11 to 2014–15. The estimated model controls for other soccer demand determinants, including economic conditions, game environment, weather conditions, game quality, fixed effects for seasons, geographical locations and team specifics. Our findings show that an improvement in one unit of game-level uncertainty measured with game-level competitive balance ratio stands to increase soccer attendance by 35 per cent, or approximately 3500 additional attendance, ceteris paribus. The outcome of this research will benefit policymakers and sport-industry stakeholders.