Simplified designs of milk-composition recording, based on information from a single monthly milking, adjusted or not for interval between milkings and for milk yield, were simulated and evaluated for 2553 ewe-test-day records belonging to 155 lactations of Churra dairy ewes. Accuracy of simplified methods was evaluated by comparing estimated trait values (fat, protein, casein and total solid yields) with those observed both in a reference plan, where the two daily milkings were recorded at weekly intervals (A1), and in the official A4 milk recording (monthly records of the two daily milkings). Trait yields per lactation were estimated and adjusted to the only milking period (days in milk 30–120). Estimates of milk component traits were less precise when monthly designs, including the A4 design, were compared with a weekly sampling of both a.m. and p.m. milkings, with fat yield being the most difficult trait to estimate. All options with one daily milking every month were more accurate when the corresponding plan was based on, or began with, the a.m. milking. Adjustment for the preceding interval between milkings or milk yield did not improve sampling accuracy. The design alternating a.m. and p.m. milkings every month, without adjustment, is suggested for recoding milk component traits in dairy ewes.