Glyphosate-tolerant (GT) alfalfa was recently deregulated and presents an option for alfalfa growers for control of weeds using glyphosate.  However, questions have arisen as to the relative yield potential of GT alfalfa varieties compared with conventional varieties, and of the effects of herbicide regime (conventional or glyphosate) on productivity.  A trial was initiated in Tulelake, CA in 2005 to compare GT and conventional alfalfa cultivars under their respective weed management systems over a 5-year period.  This is a 3-4 harvest region, and varieties are typically ranging between 3 and 5 Fall Dormancy Rating.  Conventional varieties were treated with a conventional herbicide program typical for the area and GT varieties were treated with the same conventional herbicides and, in a separate block, treated with glyphosate in a Randomized Complete Block Design with a modified split plot restriction.  Averaged over all 12 GT cultivars, the first-year alfalfa yields were 1.1 Mg/ha greater when the alfalfa was treated with glyphosate than when treated with imazamox, a conventional herbicide used during stand establishment.   In the subsequent four years when alfalfa was treated with winter-dormant herbicides, there was not a consistent difference in yield when GT cultivars were treated with glyphosate versus conventional herbicides.  However, over the 5-year stand life, alfalfa yield was 2.5 Mg/ha greater when the GT alfalfa was treated with glyphosate vs. traditional herbicides, averaged across varieties.  All varieties yielded more than Vernal, the check cultivar.  However, as a group, the yield of GT cultivars was similar to conventional cultivars when grown under the same conventional herbicide regime.   There were significant differences between individual cultivars within both the conventional and GT groups.   In addition to this test at a single location, GT varieties have been tested over 8 years in a range of California environments in standard variety trials.   Three-year average annualized yields in these imbalanced tests for GT and conventional alfalfa varieties, respectively were:  22.4(12) and 22.6(8) Mg/ha (Davis, 2004-06); 26.9(10) and 27.1(44) Mg/ha (Parlier, 2005-07); 22.8(3) and 24.2 (42) Mg/ha (Davis, 2006-08); 23.1 (10) and 24.6 (30) Mg/ha (Five Points, 2007-09); 18.2 (23) and 18.2 (22) Mg/ha (Davis, 2007-09);24.2(10) and 25.5 (47) Mg/ha (Parlier, 2008-10); 16.8(26) and 17.0(30) Mg/ha (Tulelake, CA) (number in parentheses indicates number of lines averaged).  All of these sites with the exception of Tulelake were semi-dormant to non-dormant varieties, with 6-9 cuts/year.   Although there was a numerical trend for lower yields in the GT vs Conventional varieties in the non-dormant varieties in these trials in some cases, this was not consistent, and was not apparent in the dormant group of varieties (FD 3-4).   In general, the differences between varieties within a group (GT or conventional) were much more important than the differences between the GT and conventional lines.  There was a yield advantage to the GT system in the Tulelake trial across varieties, due to crop injury in the conventional system.