HST ACS and NICMOS data are now of sufficient depth and areal coverage to place strong constraints on the formation and evolution of galaxies during the first 1-2 Gyrs of the universe. Of particular interest are galaxies at z similar to 6 since they represent the earliest epoch accessible to current high-efficiency optical instrumentation. To this end, we have been involved in the systematic construction of a large sample of 346 z similar to 6 objects from all the deepest wide-area HST data (UDF, UDF-Parallel, and GOODS fields). They have been used to construct an optimal determination of the rest-frame continuum UV LF at z similar to 6. Our LF extends to over 3 magnitudes below L*, fainter than has been done at z similar to 3. Over the interval z similar to 6 to z similar to 3, we find strong evidence for evolution in the UV LF. Though we can thus far make no strong claims on an evolution of the faint end slope, the characteristic luminosity appears to have approximately doubled over the interval z similar to 6 to z similar to 3, consistent with hierarchical expectations. Remarkably, this shift to lower luminosities extends to even higher redshifts. Using all deep J + H NICMOS observations (800 orbits in total), we have been able to demonstrate that the bright end of the LF (> 0.3L(z=3)*) is at least 5 times lower at z similar to 10 than at z similar to 4, with a similar deficit being established from our recent detections and first statistical sample of z similar to 7-8 galaxies using our UDF NICMOS data. In these precedings, we discuss what is known about the UV LF and UV luminosity density at z similar to 6-10 from current data and its evolution relative to z - 3. We also describe several exciting prospects for advance in this area over the next year. (c) 2005 Elsevier B.V. All rights reserved.