In civil litigation voir dire (jury selection) has a fundamental impact on jury composition, and by extension, is inextricably entangled with constitutional, statutory, and precedential guarantees regarding an impartial jury of the parties' peers. However, despite its deep influence on jury composition, and by extension trial outcomes, the rules for jury selection vary wildly by jurisdiction, ranging from almost no questioning of jurors at all to extensive questioning that dives into details of the underlying cases. Indeed, selection can take a week in some states and 20 minutes in others. This article involves a first of its kind empirical study that analyzes results from a national sample of over 2,000 mock jurors. The study measures the impact of jury selection across three distinct civil case presentations based on actual cases. It measures existing jury biases, explores how those biases impact juror decisions, considers whether existing methods common to voir dire sufficiently identify those biases, and analyzes whether existing methods for curing such biases are effective. It concludes that a) existing biases, many of which are prevalent in the existing jury population, fundamentally impact juror decision making - sometimes in ways that are inappropriate, b) these biases can harm both sides of civil litigation but they skew in favor of civil defendants, c) these biases will often go unidentified under existing rules so that an impartial jury is not assured, d) this problem is not cured by existing interventions, and e) jurors are largely blind to the role bias plays in their decision making. The article concludes with best practices that could be useful to attorneys, legislators, and judges.