From the Hinshaw et al WMAP paper, fits of the low-order multipoles of the CMB sky.
Here's a figure from the WMAP temperature results paper (available here) that drew some attention from the theorists in recent discussions. In part because of the provocative-sounding term attached to it: the Cosmic Axis ... of Evil!
First, what's going on in this figure? The top left figure is the familiar temperature map of the microwave background. Now the bread-and-butter of CMB work is breaking this map up into multipoles, or simple functions that each encode structure on a particular scale, and which when added together give you the original map. Reading the maps left to right, top to bottom, the first few multipoles are shown that add up to the large-scale structure of the CMB sky. In the real map, notice that there's a dark blue (cold) patch just right of the center. This sort of large-scale structure is reflected in the multipole plots; several of these low-order maps similarly have a cold peak at about this point.
When this was first done a few years back, ears pricked up because, if you squint, it looks like the l=2 and l=3 (and mayle l=5) multipoles have the same alignment. Almost like they're lined up along a cosmic axis, which you wouldn't expect if the multipoles are randomly aligned. But it's theoretically very naughty to give the Universe any kind of special direction; hence the axis of evil bit. In particular, it's hard to have a preferred cosmic axis, or vector anisotropy, without messing up the electromagnetic force in really obvious ways.
But back then it was pointed out that the supposed Cosmic Axis also lines up with the axis of the galactic coordinate system, and that would be quadruply unlikely. So it was dismissed as an artifact of not being able to perfectly subtract contamination from the galaxy -- for instance, maybe the cold patch I mentioned above isn't real. Except that now we have the 3-year WMAP data release, and it makes a strong case that this is real. So either we have a curious coincidence on our hands (just how curious is being hotly debated, but any sort of curious coincidence always makes theoretical physicists jumpy), or there's something genuinely odd about the very geometry of our universe. So far they think it's not quite curious enough that we need to seriously consider the second possibility, but be sure that they're thinking about it.