Daily increments in stem radius were determined from hourly dendrometer measurements in each of three irrigated Eucalyptus nitens and E. globulus trees. Multiple regressions determined from daily weather variables accounted for 40–50% of the variance in increment. The use of weather variables lagged by 1–2 days increased the variance explained. The diurnal variation in stem radius was resolved into three mathematically defined phases: shrinkage, recovery and increment. The positive daily net increment in stem radius, by definition, occurred in the increment phase. Average weather conditions during this phase (predominantly night-time) did not explain any more variance in increment than the average daily conditions, determined over a 24 h period. Daily increment was resolved into a rate of stem radius increase during the increment phase and the duration (hours) of that phase. Significant species by month interactions were evident with growth in summer characterised by faster rates of stem expansion over shorter time periods within each diurnal cycle. E. nitens tended to have longer increment phases in spring and autumn, and faster phase rates in autumn than E. globulus. Interactions between weather variables and cambial growth were complicated and varied over the year. The correlation between temperature and stem growth varied from positive in spring to zero or negative during summer. The data indicate a need to understand weather-by- climate interactions at the level of whole tree physiology in order to fully understand the effect of weather on cambial activity and therefore stem increment and wood properties.