[My views only, not my past or present employers']
My knowledge is out of date but I thought a lot about this about 8-10 years ago when I strongly considered getting a PhD in economics and worked as an economics research assistant at the Brookings Institution. All of the below should be caveated w/ 'as of 8-10 years ago.'
In general, +1 to dmolling's comment.
As of then, the LSE Masters in Econometrics and Mathematical Economics seemed to stand out as by far the best Econ Masters program if you wanted to go on to a PhD. Note that LSE also has other Masters in Economics programs that are more geared towards being a terminal masters and going on to work in policy as opposed to getting a PhD.
Agree with dmolling that US Masters in Economics are generally not seen as good feeders for PhD programs. Seen as low status among academic economists and geared towards getting a policy job immediately as opposed to getting a PhD. It's possible another exception is the MPAID program at Harvard, which I think makes you take most of the first year curriculum for econ PhD students (or at least used to). Make sure I'm right about this before applying.
The main other options are Masters in Statistics, applied math, or other quantitative areas. I don't know how these compare to the LSE Econometrics program but top math/stats programs probably beat US masters in econ programs. I'd guess Computer Science is also becoming a more popular background.
In general, I think the main things econ PhD programs judge you on are: 1) how many math/stats courses you've taken, how hard they were, and how well you did; 2) recommendations form top economists that the people on the admissions committee know and trust.
I was always told that a top priority was taking (at least) linear algebra, real analysis, and a couple semesters of statistics. Hopefully differential equations, too, and ideally a bunch more math than that. My guess is that taking courses like these and excelling in them is more important than whether or not you come out with a particular degree so, as dmolling says, you could just take them w/o being enrolled in a grad program. Heuristic I've gotten is that if you don't get an A in real analysis, you're probably not a good candidate for a top 6 econ program (but getting an A in real analysis is far from a guarantee; most people at those programs will have taken a lot more math than that).
General advice is that you should ask a Masters program for a list of where alums of the last several years went after finishing. For an econ Masters, if few of them went on to PhD programs you'd be happy at, assume the degree won't help your admissions prospects much.