This paper investigates whether new venture performance becomes easier to predict as the venture ages: does the fog lift? To address this question we primarily draw upon a theoretical framework, initially formulated in a managerial context by Levinthal (Adm Sci Q 36(3):397–420, 1991) that sees new venture sales as a random walk but survival being determined by the stock of available resources (proxied by size). We derive theoretical predictions that are tested with a 10-year cohort of 6579 UK new ventures in the UK. We observe that our ability to predict firm growth deteriorates in the years after entry—in terms of the selection environment, the ‘fog’ seems to thicken. However, our survival predictions improve with time—implying that the ‘fog’ does lift.