It doesn't matter how many times you test subject lines. You will never be able to reliably identify the winner without testing every time.
Here's how to do it.

Forty eight hours before each send, test three or four different subject lines against a small portion of the list.

So if you have a list of a 100,000 names, take out three groups of 1,000 names each and send each a separate test subject line.

(Don't stress about the final validity of the test. What we are looking for here are strong indicators, not a statistically significant set of research data for publication.)

Twenty four hours later, look at the report on open rates and conversion rates, if appropriate.

You will constantly be surprised by what you see. Often you'll find your favorite subject line bombing, and the one you hated delivering double the open rate.

Once you have your winning subject line, use it for the major send a few hours later.

Test three or four options for every promotional email or newsletter you send.

If that sounds like a lot of work, do the math. Let’s say that each time you test, the best-performing subject line does eight percent better than the worst. And that’s a very conservative figure. Now count up the number of times you send promotional emails and newsletters each year. Finally, do the math on revenues.

Increasing open and/or conversion rates by eight percent with every send can have a huge impact on your ROI after 12 months.