Characterising the large-scale structure and plasma properties of the inner corona is crucial to understand the source and subsequent expansion of the solar wind and related space weather effects. Here we apply a new coronal rotational tomography method, along with a method to narrow streamers and refine the density estimate, to COR2A/STEREO observations from a period near solar minimum and maximum, gaining density maps for heights between 4 and 8R . The coronal structure is highly radial at these heights, and the streamers are very narrow, in some regions only a few degrees in width. The mean densities of streamers is almost identical between solar minimum and maximum. However, streamers at solar maximum contain around 50% more total mass due to their larger area. By assuming a constant mass flux, and constraints on proton flux measured by Parker Solar Probe (PSP), we estimate an outflow speed within solar minimum streamers of 50-120 kms −1 at 4R , increasing to 90-250 kms −1 at 8R . Accelerations of around 6 ms −2 are found for streamers at a height of 4R , decreasing with height. The solar maximum slow wind shows a higher acceleration to extended distances compared to solar minimum. To satisfy the solar wind speeds measured by PSP, there must be a mean residual acceleration of around 1-2 ms −2 between 8 and 40R . Several aspects of this study strongly suggest that the coronal streamer belt density is highly variable on small scales, and that the tomography can only reveal a local spatial and temporal average. Subject headings: Sun: corona-sun: CMEs-sun: solar wind 2010), and regions in between (e.g. Kramar et al. 2009); (ii) the short-term (Morgan 2011a; Vibert et al. 2016) and long-term (Morgan & Habbal 2010; Morgan 2011b) time evolution of the corona; and (iii) the relationship between magnetic and plasma structures (Vásquez et al. 2008; Morgan & Habbal 2010; Morgan 2011a; Kramar et al. 2014). Furthermore, estimates of the 3D distribution of density is crucial for interpretation of observations by other instruments (Frazin et al. 2003), and as boundary constraints for large-scale models (Frazin et al. 2005). These methods, and their findings, are described in the review of Aschwanden (2011). Most of the literature is dedicated to descriptions and tests of methods, with application to a number of case studies. Exceptions that look at larger datasets and longer periods are Morgan & Habbal (2010) and Morgan (2011b), where more long-term studies of the streamer distribution are made, albeit based on a qualitative view of density structure. CRT of the solar maximum corona is particularly difficult, and results from these periods are rare (Butala et al. 2005; Morgan & Habbal 2010).Recently, Morgan (2019) (hereafter Paper II) developed a new CRT method based on spherical harmonics. The success of the method depends on the detailed calibration and pre-processing methods described by Morgan (2015) (hereafter Paper I). The method is restricted to heights where the coronal structure is predominantly radial, so at helio...