This article investigates sample selection bias in early-stage investment. We use comprehensive administrative data on the universe of new firm starts in Norway, allowing us to compare venture-backed firms with ex ante similar firms that do not receive venture funding. The valuation premium for venture backing is sizeable at firm birth and doubles over the first 5 years, implying a substantial upward bias in venture capital (VC) returns relative to comparable firms. In contrast, the premium for firms receiving multiple rounds of outside equity emerges only after the first year and remains significantly smaller than the VC premium throughout the firm life cycle.