In this study silent reading by adults ranging in age from 35 to 90 years was investigated. The texts to be read were printed in black on white paper with character sizes varying from 1–9 mm x-height (visual angle 0.19° to 1.67°). In order to separate age effects from visual-acuity effects, subjects with different levels of visual acuity (0.1–2.5 decimal acuity) participated in the experiment. Silent reading rate was employed as the dependent variable. Visual acuity affected reading rate most, followed by letter size. In normal-acuity subjects the variance in reading rates decreased as a function of age. Reading rates initially increased rapidly with increasing letter size, but after reaching an optimum gradually declined again as letters became larger. For the different acuity classes there appeared to be clearly optimal letter sizes, varying from 1.9 for the highest acuity group to 6 mm for the lowest acuity group at the 33 cm reading distance employed. However, the optimal reading rates of visually impaired subjects found in this study remained below those of individuals with normal acuity. This suggests that visual impairment is a more general neural phenomenon rather than merely a deficient optical image. The obtained reading-rate data could be accurately described by a theoretical model encompassing a decoding process and an integration process. It appeared that the model predictions were entirely determined by the smallest letter size at which reading is just possible with a specified visual acuity. It is concluded that both decoding and integration are dependent on visual acuity and that, in the absence of specific visual defects, ageing effects in reading can be completely explained by gradual lowering of visual acuity having its origin in central mechanisms.