The objective of this study is to solve the debate from practitioners that which of mutual funds with high or low net asset value per share (NAV) should be selected by investors. This study examines the relationships between NAVs and both the further long- and short-term performance and the performance persistence for Taiwan’s equity mutual funds from 1995 to 2008. The study methods include the equality tests of various performance indexes on sub-groups sorted by NAVs and the multivariate regression analyses with panel data. The findings are as follows: after controlling past performance and other factors affecting returns, there is a strong explanatory power for NAVs to future performance, that is, regardless of short- or long-run returns, there is a negative relation between NAVs and future performance. The further annual and three-year returns will reverse from the past returns over the same long periods, and the reversal behaviors dissipate for five-year returns, without positive persistence. NAVs do not affect the relationship between past and future performance for annual and five-year, respectively, but there is an effect on the degree of negative relation for three-year performance. Since the NAV is the data relatively available to investors, this study suggests that investors should make investment decisions according to the investment horizon length, three-year returns and NAVs. If the horizon is one year, the funds with relatively low NAVs and lower past performance should be selected. For a three-year investment, investors selecting the funds with the lower past performance and higher NAVs (not lower than $24.66) will yield the larger degree of performance progression; if selecting funds with higher past performance, then the funds with moderate NAV level (6.09-18.64) can be invested to mitigate the extent of performance reduction. Finally, for a five-year investment horizon, the finds with relatively low NAVs should be invested and do not consider their past performance. This result answers the question of practitioners and fills the research gap about funds’ performance persistence.