Well, there are others like RIFT, G-RIF, SURF, PCA-SIFT that you can read about here: http://is.gd/a1Kb40
As of 2005 (SIFT was first published in 1999), SIFT and SIFT-like techniques apparently outperformed others within limits, according to Mikolajczyk, K., and Schmid, C., "A performance evaluation of local descriptors", IEEE Transactions on Pattern Analysis and Machine Intelligence, 10, 27, pp 1615--1630, 2005. You can see a summary here: http://is.gd/xN3Kj6
Speed Up Robust Features (SURF) has become very popular since 2006 and updated in 2008: http://en.wikipedia.org/wiki/SURF.
Local Energy based Shape Histogram (LESH) is another "new" one as of 2008: http://en.wikipedia.org/wiki/LESH
I don't know about "best" as I don't don't know all of the trade-offs or your application. I've had SIFT used on 3D shape-from-motion projects that worded very well, and SURF-128 (http://www.vision.ee.ethz.ch/~surf/eccv06.pdf) on pose estimation for navigation that worked better than SIFT under those specific conditions in the project.