Cotton is a perennial plant with an indeterminate growth pattern that is typically produced like an annual, but requires proper management to effectively produce high yields and good fiber quality in a thermally limited environment like the northern Texas High Plains. In 2007 and 2008, we investigated the effect of irrigation scheduling/control method and amount on cotton (Gossypium hirsutum L) yield and water use efficiency. Methods were automatic irrigation scheduling and control of a center pivot system, and manually scheduled irrigation to replenish soil-water to field capacity. Cotton was irrigated with LEPA (low energy, precision application) drag socks in furrow dikes; three blocks were irrigated manually and three were irrigated automatically. Six replicates of the manual and automatic irrigation treatments were included in the randomized block design. Manual irrigations were based on the weekly replenishment of soil-water to field capacity in the top 1.5 m of the soil profile and included a fully irrigated treatment (I-100), and treatments receiving 67% (I-67) and 33% (I-33) of the I-100 amount, plus a non-irrigated treatment (I-0). Automatic irrigations were triggered using a time temperature threshold (TTT) algorithm, which was designated as the I-100 treatment, and treatments receiving 67%, 33%, and 0% of that amount (I-67, I-33 and I-0, respectively). In 2007, overall mean lint yields (102.3 and 101.6g m(-2), manual and automatic, respectively) were not significantly different. Similarly, yields were not significantly different across automatic and manual treatments in the same treatment level, with the exception of the I-67 treatment where the manual treatment yields were 11% greater. In 2008, the mean yields were 70% less than those in 2007 for both methods of irrigation (30.3 and 30.9 g m(-2). manual and automatic, respectively) due to harsh climatic conditions at emergence and heavy rainfall and cooler temperatures in the month of August. Yields from the automatically irrigated plots in the I-100 and I-67 treatments, however, were significantly greater than yields from the corresponding manually irrigated plots; though there was no significant difference between yields in the drier treatments (I-33 and I-0) plots. These results indicate that the M algorithm is a promising method for auto-irrigation scheduling of short season cotton in an arid region. However, further studies are essential to demonstrate consistent positive outcomes. Published by Elsevier B.V.