Determining causes of poor reading ability is an important step in trying to ameliorate reading performance in low-vision patients. One important parameter is word acuity. The principal aim of the current study is to develop a method to reliably measure acuities for isolated lowercase letters and words of differing length that can be used to test low-vision patients. Using isolated stimuli means that testing is relatively free of potential crowding and/or distracting attentional effects from surrounding words, it is unambiguous which stimulus subjects are trying to read and response times can be recorded for each stimulus. Across a series of experiments, subjects with normal vision were asked to read isolated lowercase single letters and lowercase words of 4, 7 and 10 letters, in separate tests. Acuities for uppercase Sloan letters were also measured to provide a reference, as they are commonly used to measure visual acuity. Each test was based upon the design principles and scoring procedures used in the Bailey–Lovie and ETDRS charts. Acuities for uppercase Sloan letters were found to be equivalent whether measured using ETDRS charts or the computer-based method. Measurement of acuities for lowercase single letters and lowercase words of 4, 7 and 10 letters had a reliability that was no worse than acuities for uppercase Sloan letters. Lowercase word acuities were essentially independent of word length. Acuities for single lowercase letters and lowercase words were slightly better than uppercase Sloan letters acuity. Optimal processing of lowercase single letters and 4-, 7- and 10-letter words occurred at character sizes that were at least 0.2–0.40 log MAR above acuity threshold, i.e. between 1.5 and 3 times threshold acuity for that particular stimulus. In general, critical character sizes appear similar across word lengths as progressive increases or decreases in these values were not observed as a function of the number of letters in the stimulus. We conclude that a computer-based method of stimulus presentation can be used to obtain highly repeatable measures of acuity for lowercase single letters and lowercase words in normal vision.