The recent calculations of atomic data for ions of astrophysical interest are reviewed with a focus on work performed in Cambridge. The calculations have been benchmarked against high-resolution laboratory and astrophysical spectra. A framework for assessing uncertainties in atomic data has also been developed. Long-standing discrepancies in predicted spectral line intensities have been resolved, and a significant number of levels in coronal ions have finally been identified, improving the modelling of the extreme-ultraviolet and soft X-ray spectral regions. Recent improvements based on collisional-radiative modelling are presented. They are relevant for the modelling of satellite lines in the X-rays and for solving the long-standing problems in the chromosphere-corona transition in stellar atmospheres.