Aim The British Isles Network of Congenital Anomaly Registers (BINOCAR) provides surveillance of congenital anomalies in England and Wales.1 The aim of this study is to investigate if there is regional variation in 11 major congenital anomaly subgroups between five registers.

Methods All cases of congenital anomalies belonging to 11 major subgroups were extracted from five registers to calculate birth prevalence rates for 2005-2009. Differences between the registers for each congenital anomaly subgroup were assessed using chi-squared tests and regional variation depicted using maps.

Results In 2005-2009, the overall prevalence of congenital anomalies was 230 per 10,000 births [95% CI: 227–233]. The regional reported prevalence of congenital anomalies ranged from 188 per 10,000 births [95% CI: 181-195] in Thames Valley to 331 per 10,000 births in Wales [95% CI: 322-339]. These are both significantly different to the reported prevalence for the five registers together.

There was significant regional variation in all major congenital anomaly subgroups except for abdominal wall defects. Wales had the highest reported prevalence in all subgroups except for chromosomal anomalies. The East Midlands & South Yorkshire and Thames Valley had the lowest reported prevalence in all but two subgroups.

Conclusion There is regional variation in the reported prevalence of ten major congenital anomaly subgroups. This variation requires investigation to determine any additional influences other than ascertainment and the underlying true prevalence which may vary regionally.