Performance records of 2,034 beef bulls tested at the Ellerslie Test Station, Alberta, Canada from 1974 to 1987 were analyzed to identify important factors that affect growth of bulls in a test station and to estimate the contribution of herd of origin to phenotypic variances of growth traits. Compared with herd of origin, age of dam, start of test age, and weight of bull were not of practical importance. Herd of origin accounted for a large fraction of phenotypic variance of weight, decreasing from 52 to 35% and from 37 to 30% from start to end of test in groups of small and large breeds, respectively. The effect of herd of origin on periodic and cumulative ADG decreased as the test advanced. The least affected periods were either 28 to 112 d or 56 to 112 d of test. However, the effect increased during the 112- to 140-d period. The effect of herd of origin on periodic relative growth rate was large and relatively stable up to 84 d, decreased during the 84- to 112-d period, and increased during the 112- to 140-d period. The effect of herd of origin on cumulative relative growth rate showed an increasing trend. Because ADG was highly influenced by herd of origin during the first 4 wk of test and least affected during the 28- to 112-d period or the 56- to 112-d period, a 56-d adjustment period followed by an 84-d test period for accurate evaluation of beef bulls for growth potential seems appropriate.