A Rhinolophus ferrumequinum maternity roost in Pembrokeshire, West Wales, U.K. was monitored from 1994 to 2023, as part of the U.K. national survey between days 92–112 (1 April = day 1). The onset of births, determined from pup forearm lengths, occurred earlier than anticipated in five years from days 63–77, in 1997, 2005, 2007, 2011 and 2014. In those five years between 0.5–20.6% of the exit count would have been pups 28–35 days old, which were nearly adult size and able to fly outside the roost (9.5 ± 8.5%, 0 ± SD). Since 2014 the population exceeded 650 adult bats and there was insufficient time to measure pup forearm lengths whilst adults were out of the roost so alternative methods of estimating birth timing were devised. Nightly external activity periods at the maternity roost identified the date when births commenced and presumed foraging activity changed from pre-partum, when bats were absent from dusk to dawn, to post-partum when mothers returned to feed their pups. Changes in the onset of early births were statistically related to changes in the integrated air temperature, defined as the accumulated temperature above a threshold in April of 7°C (degree days). A significant (P < 0.01) polynomial equation explained a moderate proportion (R² = 33.6%) of the non-linear increase in the integrated air temperature over time. The probability of an early birth onset (P) was predicted by a significant (P = 0.014) binary logistic model. The odds of an early birth onset, relative to a later birth onset, was elevated by 1.04 (95% CI = 1.01, 1.08) for every one degree day increase in integrated air temperature. The probability of an early birth onset increased from P = 0.107 when the integrated air temperature was 38.7 degree days up to P = 0.942 when the integrated air temperature was 152.0 degree days. At the principal U.K. maternity roosts integrated air temperatures in April were in the same ranges annually from 2011 to 2023. Therefore, monitoring of the integrated air temperature in April provides a simple noninvasive method to predict the extent and pace of the shift in birth timing of R. ferrumequinum in future associated with climate change.