Venture capitalists (VCs) are considered experts in identifying high-potential new ventures—gazelles. VC-backed ventures survive at a much higher rate than those ventures backed by other sources (Kunkel and Hofer 1991; Sandberg 1986; Timmons 1994). Thus, the VC decision process has received tremendous attention within the entrepreneurship literature. Nonetheless, VC-backed firms still fail at a surprisingly high rate (20%). Moreover, another 20% of the VC's portfolio fails to provide any return to the VC. Therefore, there is room for improvement in the VC investment process. The three staged investment process often begins with venture screening. First, VCs screen the hundreds of proposals they receive to assess which deserve further consideration. Those ventures that survive the initial stage are then subjected to extensive due diligence. Finally, the VC and entrepreneur negotiate terms of the investment. Considering the amount of time that due diligence and negotiation of terms may take, it is imperative that VCs minimize their efforts during screening so that only those ventures with the most potential proceed to the next stage. Yet, at the same time, the screening process should also be careful not to eliminate gazelles prematurely. VCs are in a quandary. How can they efficiently screen venture proposals without unduly rejecting high potential investments? The answer may be to use actuarial decision aides to assist in the screening process. Actuarial decision aides are models that decompose a decision into component parts (or cues) and recombine those cues to predict the potential outcome. For example, an actuarial model about the VC decision might decompose a venture proposal into decisions about the entrepreneurial team, the product, the market, etc. The sub-component decisions are than recombined to reach an overall assessment of the venture's potential. Such models have been developed in a number of decision domains (e.g., bank lending, psychological evaluations, etc.) and been found to be very robust. Specifically, these models often outperform the very experts that they are meant to mimic. The current study had 53 practicing VCs participate in a policy capturing experiment. The participants examined 50 ventures and judged each venture's success potential; would the venture ultimately succeed or fail. Likewise, identical information about each venture was input into two different types of actuarial models. One actuarial model—a bootstrap model—used information factors that VCs had identified as being most important to making a good investment decision. The second actuarial model was derived by Roure and Keeley (1990). The Roure and Keeley model best distinguished between success and failure in a study of 36 high-technology ventures. The bootstrap model outperformed all but one participating VC (he achieved the same accuracy rate as the bootstrap model). The Roure and Keely model, although less successful than the bootstrap model, outperformed over half of the participating VCs. The implications of this study are that properly developed actuarial models may be successful screening decision aides. The success of the actuarial models may be attributed to their consistency across different proposals and time. The models always weight the information cues the same. VCs, as are all human decision makers, may often be biased by differing salient information cues that cause them to misinterpret or ignore other important cues. For example, a VC may overlook product weaknesses if (s)he is familiar with the entrepreneur putting forth a particular proposal. Although the current study developed a generalized actuarial model, each VC firm could create screening models that fit it's particular decision criteria. The models could then be used by junior associates or lower level employees to perform an initial screen of received venture proposals thereby freeing senior associates' time.