Does our ability to predict the performance of new ventures improve in the years after start-up? We investigate the growth and survival of 6247 new ventures that are tracked using the customer records at Barclays Bank. We put forward Gambler’s Ruin as a simple theory for understanding new venture growth and survival. Gambler’s Ruin predicts that the R2 remains low for growth rate regressions, but that the R2 increases in the years since start-up for survival regressions. The Nagelkerke R2 obtained from growth rate regressions decreases significantly in the years after start-up, which suggests that the fog gets thicker with respect to growth. When we focus only on firms surviving until the end of the period, however, there is no visible change in the R2 over time. In contrast, the Nagelkerke R2 of survival regressions increases in the years after start-up. Interestingly, a blip in year 5 suggests that macro-economic factors may have a strong effect on the amount of ‘fog’.