We sought to determine the normal within-sheep variability of bronchoalveolar cellular parameters and to infer the numbers of sheep required to detect a defined significant change within the same. We further sought to examine the effect of repeated bronchoalveolar lavage (BAL) on those parameters. Neither the total cell number nor the absolute numbers of any given cell type changed significantly following repeated BAL. Additionally, there was no significant change in percentage differential cell populations over time in randomly sampled lobes. However for lobes sampled repeatedly, a significant change in the percentage of lymphocytes was detected. Although the percentages of neutrophils in repeatedly and randomly sampled lobes were significantly correlated, the percentage found in the former tended to be greater and more variable than in the latter. As a consequence, larger group sizes are required to detect given changes in neutrophil percentages in repeatedly sampled lobes over time when compared with detecting equivalent changes in randomly selected lobes.