Weed management, training time, and irrigation practices were evaluated from 2013 to 2014 in a mature field of trailing blackberry (Rubus L. subgenus Rubus Watson) established in western Oregon. The field was planted in 2010 and certified organic in 2012, before the first harvest season. Treatments included two cultivars (Marion and Black Diamond), three weed management practices [nonweeded, hand-weeded or bare soil, and weed mat (black landscape fabric)], two irrigation strategies (irrigation throughout the growing season and no postharvest irrigation), and two primocane training dates (August and February). When averaged over the other treatments, ‘Marion’ and ‘Black Diamond’ had similar yields in both years. However, the presence of weeds reduced vegetative growth and yield, especially in ‘Black Diamond’, while weed mat increased growth and yield over hand-weeded plots by 13%. Withholding irrigation after harvest reduced water use by an average of 44% each year without adversely affecting yield in either cultivar. The effects of training time were primarily seen in 2014 after a cold winter. August-trained ‘Marion’ plants had more cold damage than February-trained plants and, consequently, had fewer and shorter canes, less biomass, fewer nodes, and 1 kg/plant less yield than February-trained plants. ‘Black Diamond’ was cold hardier than ‘Marion’, but was more readily infested by raspberry crown borer (Pennisetia marginata Harris). As the planting reached maturity, yields in the best performing organic production systems (both cultivars under weed mat and ‘Marion’ that was February-trained) averaged 11 and 9 t·ha−1, for ‘Black Diamond’ and ‘Marion’ respectively, similar to what would be expected in conventional production.