Abstract Large, naturally weathered mineral fragments are often ground and sieved to obtain samples for dissolution studies. If the fragments are ground to much smaller dimensions, the samples are normally assumed to contain one type of (fresh) surface only. A model has been developed to test the validity of this assumption. The model describes the surface roughness factor of ground mineral material as a function of grain size and can be used to: 1. (1) estimate the roughness factors of the freshly created and the naturally weathered surfaces; 2. (2) estimate the ratio of fresh to total surface area for the ground samples, both for geometric and for actual surfaces; and 3. (3) check the internal consistency of surface area measurements. Literature data were evaluated for intensively ground size fractions of nine different naturally weathered feldspars. Roughness factors of freshly created surfaces ranged from 2.5–11. Roughness factors of naturally weathered surfaces ranged from 130–2600, which is much higher than is generally recognized for feldspars. Comparison with surface roughness estimates from Scanning Electron Microscopy strongly suggests that etch pit formation plays a minor role in the increase in actual surface area during natural weathering. Instead, virtually all increase in surface area must be attributed to the formation of internal surface structures like micropores. The model also showed that for these ground samples, the assumption of one type of (fresh) surface is approximately correct for the geometric surfaces. For the actual (BET) surfaces, ratios of fresh to total surface area varied over almost the entire range from 0 to 1. This demonstrates that, even after intensive grinding, samples from large, naturally weathered mineral fragments can still contain substantial proportions of weathered BET surface area. Thus, previous dissolution studies in terms of fresh BET surface only may have been misinterpreted.