Livestock produce 10% of the total CO2-equivalent greenhouse gases in Australia, predominantly as methane from rumen fermentation. Genetic selection has the potential to reduce emissions and be adopted in Australian grazing systems. Developing a breeding objective for reduced methane emissions requires information about heritability, genetic relationships, when best to measure the trait and knowledge of the annual production of methane. Among- and within-animal variation in methane production, methane yield and associated traits were investigated, so as to determine the optimal time of measurement and the relationship between that measurement and the total production of methane. The present study measured 96 ewes for methane production, liveweight, feed intake, rumen volume and components, and volatile fatty acid (VFA) production and composition. Measurements were recorded at three ages and different physiological states, including growing (12 months), dry and pregnant (21 months) and dry (non-pregnant, non-lactating; 28 months of age). The single biggest determinant of methane production was feed intake, but there were additional effects of age, proportion of propionate to (acetate+butyrate) in rumen VFA, total VFA concentration and CO2 flux. Rumen volume and pregnancy status also significantly affected methane production. Methane production, CO2 flux, liveweight, feed intake and rumen volume had high repeatability (>65%), but repeatability of methane yield and VFA traits were low (<20%). There were no interactions between sire and age (or pregnancy status) for methane traits. This suggests that methane could be measured at any time in the production cycle. However, because MY is reduced during pregnancy, it might be best to measure methane traits in dry ewes (neither pregnant nor lactating).