Which characteristics of a start-up move a venture capitalist from interest to investment? Five decades of research have produced a rich but fragmented answer, scattered across disciplines, methods and outcome definitions. We systematically review 75 studies published between 1984 and mid-2025 and organize their evidence in three steps. First, we inductively derive a taxonomy of 29 evaluation criteria in eight clusters spanning human capital, market, technological, financial, organizational, relational, institutional and process factors. Second, a weight analysis computed over two nested evidence bases separates the criteria investors say matter (perceived importance) from those that statistically predict funding outcomes (revealed importance). Third, a random-effects meta-analysis pools 122 effect sizes drawn from 40 studies. Ten criteria satisfy a dual retention criterion, qualifying as best predictors in the inferential-only weight analysis while returning significant positive pooled effects: founder/team experience, founder/team education, team composition and balance, market size and growth, competition intensity, product/technology uniqueness, intellectual property, business model and strategy, social network ties, and the use of formal decision aids. The largest venture-level pooled effects concern market size (beta = 0.41) and product uniqueness (beta = 0.33), while the strongest cumulative predictor in the corpus is investor-side: formal scoring models and algorithmic decision aids (beta = 0.42). Founder gender pools to a null overall yet masks documented context-specific biases. We map the criteria onto the staged investment process, show where perceived and revealed importance diverge, and derive an agenda for research on structured, bias-aware decision support in entrepreneurial finance.

JEL Classification: G24, L26, M13, D81