Abstract Global irradiance spectra vary with location, different viewing angles and times of day, depending on the fraction of direct and diffuse irradiance. Owing to big differences in spectral responses, PV module technologies might therefore show a differing behaviour with varying orientation and tilt angles. The purpose of this work is to verify the thesis, that thin film modules are – due to their spectral response – more suitable for horizontal orientation than crystalline. Diffuse irradiation (except from circumsolar radiation) can be captured best by a horizontal surface and consists to a greater fraction of short wavelengths than direct irradiation. At the same time thin film modules primarily absorb photons of short wavelengths and could therefore be better suited for horizontal application. Based on the semi-empirical spectral model Sedes2 and quantum efficiency data, a model has been developed to analyse differences in optimum orientation of several PV module technologies. In a first step, hourly global irradiance spectra are generated from a 1 year dataset of hourly climate data derived from long-term averages by the Meteonorm database for two sites in different climes. Based on this, average photocurrent densities are computed for each technology and for a matrix of different orientation and tilt angles using quantum efficiency data. Normalised to their maximum, the photocurrent densities are compared between the technologies. The results we obtained show, that for each site the maximum relative photocurrent densities are located at about the same orientation for all technologies, i.e. the optimum orientation is the same. At horizontal orientation, thin film modules show a slightly higher value of normalised average photocurrent densities than monocrystalline modules. Yet, for a whole year this advantage lies below 1% for both sites.