Intra-tooth sequential analysis of enamel δ18O is currently used to investigate birth seasonality in past animal populations, offering new insights into seasonal availability of animal resources, herd management and seasonality of site occupation. Reference data sets are still required to address two major difficulties: (1) that inter-individual variability in the record of the seasonal cycle is affected by tooth size; and (2) that the season of birth cannot be directly estimated from the timing of tooth growth, because of a delay in enamel mineralization. We present a data set acquired on the lower second molar of 10 modern sheep from Rousay (Orkney) born within a few weeks of each other in April/May and submitted to the same environmental conditions until death. All sheep have recorded a sinusoidal pattern of δ18O variation spanning approximately a year. From the difference between the expected and the measured time sequence, the delay of enamel mineralization is estimated to be 5-6 months. The data set is then described using a model mainly based on a cosine function. The period, corresponding to the length of the M2 crown formed over a year, averaged 35.8 mm. A very slight variation of tooth growth rate with time and no attenuation of the isotopic signal towards the cervical margin of the crown could be detected in this data set. The lowest δ18O values, corresponding to the sheep's first winter, were tracked at a distance from the enamel/root junction that varied between 23.0 and 30.3 mm (xmin mean = 27.6 mm); the highest δ18O values, corresponding to the sheep's second summer, were between 6.3 and 11.6 mm (xmax mean = 9.9 mm). Most of the variability can be attributed to tooth size. When normalized on the period, xmin and xmax are 0.28 ( ± 0.05) and 0.78 ( ± 0.05) on average, meaning that the Rousay sheep have recorded the minimum and maximum δ18O values on average at 78% and 28%, respectively, of the end of the periodic cycle recorded in the second molar. [ABSTRACT FROM AUTHOR]