Lightning and associated weather (e.g., hail, severe wind, tornado, and heavy rain) can cause severe impacts; therefore, it is important to understand how frequently thunderstorms occur and how the frequency might be changing with time. Published studies until now have been limited to relatively short time periods or extrapolated from low‐resolution surface data. This article presents the methods and results to combine long‐range lightning location network data (precise but inconsistent) with surface observations (less precise but more consistent) to generate a more consistent, more precise gridded days of thunder (DoTs) dataset for the United Kingdom. Both a 30‐year average (1990–2019) and rolling 10‐year averages are evaluated. For the 30‐year averages there is a maximum of ∼20$$ 20 $$ DoTs per year within the domain (over France) with a minimum of between 0 and 1 DoT per year (over the sea), with most of the UK land area having ∼10$$ 10 $$ DoTs per year. The summer shows the peak in DoTs relative to the other seasons. The 10‐year moving averages at either end of the 30 years allow for the trend in DoTs to be assessed. There has been a small (∼1$$ 1 $$ day) reduction in the average DoTs. There is a noteworthy north–south divide (at ∼54∘$$ 5{4}^{\circ } $$ N) with a small increase in the number of DoTs per year in the north of the UK and a small reduction in the south. The trends in 10‐year averages of DoTs are analysed for different regions of the United Kingdom (north land, south land, sea, land) over the four seasons. North land summer shows a small (∼0.5$$ 0.5 $$) increase in DoTs. All the other areas show a decrease.