公司法的累積投票制， 一直是教學與理解上的困擾點。公司法中的累積投票，要求將每股股數乘以應選董事席位，以作為各股東投票權的總數。然而，為何乘以應選董事席位會產生累積投票制所宣稱的「保障少數股東董事會代表席次」的效果，卻一直是個謎\r 團。 本文以帶入數例實際演算案例的方式，推算各種設定下的投票模型，發現現行法中將投票權數依照應選董事席位予以膨脹，其實跟一般一股一票的情形，並沒有明顯不同；反而是由於投票權數的擴張，加強了分裂的風險與計算的困擾。因而，在這樣的懷疑下， 本文藉由美國早期文獻交互查核，重新確認美國法上所謂的累積投票制的操作邏輯，以及在探討該制度移植到臺灣來後所出現的矛盾情況；最後並維持「每股股數乘以應選董事席位」作為投票數的作法並無實際效用的結論。 Cumulative voting has been a point of confusion in corporate law, from both the logical and pedagogical perspectives. According to the current rule in Taiwan, cumulative voting is required for all directorial elections, which need to multiply the number of shares to the seats of director to be elected and get the total votes for each shareholder. However, how the unique way of voting creates the promised effect of “providing better protection to minority shareholders, by giving them a better chance of having their representation on board,” is still unclear. This study first uses hypothetical models with sets of variables to simulate scenarios that might happen in real-world situations to see how the cumulative voting system works. Surprisingly, these simulations do not produce results corresponding to the assertion of “to better protect minority shareholders.” Even worse, multiplying the number of shares to the seats to be elected indeed makes the traditional one-share-onevote system more complex and imposes an extra burden in implementing voting strategy. To answer why the current system persists, this article goes back to the early literature in the United States. By cross-checking the historical roots and later developments of cumulative voting, the study arrives at an explanation that both answers to the current system and at the same time affirms the main thesis of this study: to multiply shares with seats to be elected is only a useless tail in its historical development; the key to better protection of minority shareholders is the way an election is held, not the number of votes issued.