Ten amblyopes had their visual acuity measured once a week for 6 weeks using a Snellen chart (whole chart and single line) and a newly developed E chart was designed specifically for testing the acuity of amblyopes. For each test at each trial, the subject's acuity responses were graphically plotted and were fitted by a least-squares straight line to the central-most data. The reliability (repeatability) of the 50% threshold acuity was compared for the 3 acuity tests. The new E chart was the most reliable and the Snellen single-line test was the least reliable. These results have complications for quantifying acuity improvement during and following amblyopia therapy. The slope of the fitted acuity line may have value in assessing the prognosis for treatment.