Two experiments are described. In experiment 1, penned sheep were fed oat grain, hay and molasses-urea block prepared with added chromic oxide (Cr2O3). Recoveries of the marker calculated from measured intakes and faecal outputs were: oats 98.0 � 5.1%, hay 97.5 � 3.2%, molasses-urea block 89.2 � 11.5%. The lower and more variable recoveries of marker from the block were attributed to uneven mixing of the marker. In experiment 2, three groups of 15 mature Corriedale wethers were allowed to graze freely, and supplements of oats, hay or molasses-urea blocks, prepared with added chromic oxide, were offered to each group in rotation during three 4-week periods. During the first two weeks of each period the group offered molasses-urea block was confined in a small yard with a block, and fed hay, to accustom the sheep to the block and to ensure that all sheep in the group were licking it. Total faecal collections were made during 14 days from the time of commencement of feeding of the prepared supplements, and the intakes of the supplements were calculated from the total amounts of marker excreted. The mean intakes over three periods (uncorrected for Cr2O3 recoveries) were : oats 1851 � 434, hay 3326 � 1014, molasses-urea block 11 7 � 1 6 8 g/head week-1. Intakes by individuals ranged from 1086 to 2679 g week-1 for oats, from 1508 to 6048 g week-1 for hay, and from 0 to 732 g week-1 for molasses-urea block. Of the 45 sheep used, 8 sheep did not consume any of the block within the measurement period, despite having consumed some block in the preceding 2-week period