Despite playing a crucial role in regulating Earth's radiative balance and driving the hydrological cycle, clouds remain inadequately characterised by both models and observations due to their substantial spatiotemporal variability. In tropical urban environments, cloud characteristics are further complicated by anthropogenic heat, abundance of aerosols, and local atmospheric dynamics. The present study characterises the clouds over an Urban region, Hyderabad, India, by studying the seasonal and diurnal variability of cloud base height (CBH) using long-term ceilometer observations (July 2019–March 2025). The monsoon season had the highest cloud occurrence frequency (70.5 %), followed by postmonsoon (28.1 %), premonsoon (12.5 %), and winter (7.8 %). Single-layer clouds dominated (≈ 67.8 %) irrespective of seasons, and the multi-layer clouds were more prevalent during the monsoon. Most clouds occurred below 2 km, with noticeable bimodal distributions during the monsoon and premonsoon seasons. While the primary mode at 1–2 km is closely linked to the planetary boundary layer (PBL) height, the second prominent cloud vertical mode is associated with the 0 °C isotherm level. The diurnal variation of primary cloud mode closely follows the PBL evolution, especially during the monsoon and postmonsoon seasons. The cloud occurrence frequency also shows a strong diurnal cycle with high occurrence of clouds during the daytime in all the seasons, probably associated with daytime heating by insolation. The efficacy of reanalysis datasets in producing accurate cloud characteristics is assessed by comparing ERA5 products with long-term statistics obtained in the present study. It is noted that ERA5-derived CBH overestimates (underestimates) low (high)-level clouds.