We evaluate the impact of Spire RO observations in the presence of other conventional and satellite observations, typically assimilated in weather forecasting centers. We apply an ensemble-variational data assimilation method and a global non-hydrostatic forecast model. Our control experiment, which is run without assimilation of Spire RO data, provides a baseline. Experiments with assimilation of Spire RO observations provide an insight on how much forecast improvements can be obtained due to assimilation of Spire RO data. We focus on the impact of RO observation errors, number of RO profiles and horizontal and vertical data density. As a result, we provide a guidance on how to make the best use of Spire RO observations towards maximum forecast improvements.