Temperament scores were recorded by two operators on a herd of Bos taurus beef cattle, including Angus and Hereford controls and various crossbred groups. The herd consisted of 765 cows, 653 calves at foot (average age 2 months), and 250 yearling heifers. Recording of temperament in the yards was carried out at weighing time in November 1982 using a 1-8 scale, and immediately afterwards when the herd was drafted for natural mating using a 1-6 scale, with higher scores indicating more difficulty experienced by stockmen in carrying out the routine weighing and drafting operations. In addition, a calving temperament score was recorded on 2121 cows calving in 1981-90 using a 0-5 scale. Both scores in the yards differed significantly among cow breed groups (P < 0.001). The range of means was 1.73 units (1.86 phenotypic standard deviations: σ) for the weighing score (Score 1) and 1.57 units (1.29σ) for the drafting score (Score 2). Corresponding data for yearling breed groups were 1.11σ for Score 1 (P < 0.001) and 0.49σ for Score 2 (not significant), and for calf breed groups 0.99σ for Score 1 (P < 0.001) and 0.90σ for Score 2 (P < 0.01). The heritability for average cow score in the yards (untransformed scale) was 0.22 ±0. (based on 176 sires), for average yearling score was 0.32 ± 0.24 (47 sire groups), and for average calf score was 0.23 ± 0.12 (53 different sire groups). Transforming data to a logarithmic scale made little difference to these estimates. Cow's calving temperament score had a heritability of 0.09 ± 0.03 (486 sires) and a repeatability of 0.20 ± 0.02. The correlation of average weighing/drafting scores for cow-calf pairs was 0.27, or 0.14 after adjustment for breed group. Correlations among breed groups were 0.49 for Scores 1 and 2 from cows, 0.65 for mean scores from cows and yearlings, 0.79 for mean scores from yearlings and calves, and 0.54 for mean scores from cows and calves. It was concluded that there were generally significant differences among breeds for temperament scores, and that breeds ranked similarly over the various age groups. Cow differences within breed were repeatable, but heritable effects were generally small.