Atmospheric gravity waves generated downstream by the orography in a stratified airflow are known as lee waves.In the present study, such mesoscale patterns have been detected, over water and in clear sky conditions, using the Advanced Infra-Red WAter Vapour Estimator (AIRWAVE) Total Column Water Vapour (TCWV) dataset, which contains about 20-year day-night products, obtained from the thermal infra-red measurements of the Along Track Scanning Radiometer (ATSR) instrument series. The good accuracy of such data, along with the native 1×1 km 2 spatial resolution, allows the investigation 5 of small scale features as the lee waves. In this work, we focused on the Mediterranean region, the largest semi-enclosed basin on the Earth. The peculiarities of this area, which is characterized by complex orography and rough coastlines, lead indeed to a possible development of these structures both over land and over sea. We developed an automatic tool for the rapid detection of areas with high probability of lee waves occurrence, exploiting the TCWV variability in spatial regions 0.15 • ×0.15 • wide.Through this analysis, several occurrences of structures connected with lee waves have been observed. The waves are detected 10 in spring, fall and summer seasons, with TCWV values usually falling in the range from 15 to 35 kg m −2 . In this article we describe some cases over the Central (Italy) and the Eastern Mediterranean basin (Greece, Turkey, Cyprus). We compared a case of perturbed AIRWAVE TCWV fields due to lee waves occurred over the Tyrrhenian Sea on 18 July 1997 with the sea surface winds from the Synthetic Aperture Radar (SAR), which sounded the same geographical area, finding a good agreement.Another case has been investigated in detail: on 2 August 2002 the Aegean sea region was almost simultaneously sounded by 15 both ATSR-2 and AATSR instruments. The AIRWAVE TCWV fields derived from the two sensors were successfully compared with the vertically integrated water vapour content simulated with the Weather Research and Forecasting (WRF) numerical model for the same time period, confirming our findings. Wave parameters such as amplitude, wavelength and phase, are described through the use of the "Morlet" Continuous Wavelet Transformation (CWT). The performed analysis derived typical wavelengths from 6 to 8 km and amplitude that may extend up to 20 kg m −2 .