I frequently get asked about what Bible translation I recommend. In fact, I get asked so frequently that I thought I’d briefly post on it.
The first thing I usually say is that the best Bible translation is the one you’ll read faithfully. I am far more concerned with that than staking a position on translation philosophy. I’m even willing to make allowance for paraphrases in this regard, though I really dislike them. You ought to be reading some version with consistency, though.
Second, I always point out that there is no one Bible translation that is consistently superior to all others. (Though paraphrases are consistently unfaithful to the text, but see my caveat above). All translations have problems; they all take liberties; they all have strengths. If you are interested in comparing and analyzing Bible translations, I recommend the Better Bibles blog.
Third, I recommend that everyone read from more than one translation. It’s a good idea to become acquainted with the basic differences in approaches to translating the Bible. I speak here of “dynamic equivalence” and “formal equivalence” (usually referred to as “literal translation”). I prefer formal equivalence, but I recommend reading from at least one translation that follows each approach. The above link contains a listing of how the versions stack up (at least for the writer of that article).
Fourth, you should pick a translation that is textually up-to-date. For example, I want a Bible that adopts readings in its running text from the Dead Sea Scrolls where they are demonstrably superior to the Masoretic Text. My test case for this is Deuteronomy 32:8 and Deuteronomy 32:43. The former should read “sons of God” (ESV; cp. “gods” in NRSV), or something like “heavenly beings” (NET Bible) or “heavenly court” (NLT) instead of “sons of Israel.” Verse 43 should read “bow down to Him, all you gods” (ESV, NRSV) or something akin to it like NLT’s “let all God’s angels worship him.” The preface of the particular version will alert you to such textual issues.
Hope this was informative in some way.