OBJECTIVE The purpose of the study is to evaluate the monocular and binocular outcome of three types of "penalization" (blurring of the sound eye) treatment of amblyopia: traditional full-time atropine or optical penalization and a new intermittent atropine regimen involving atropine instillation 1 to 3 days a week. METHODS The study design was a retrospective study. METHODS A total of 163 patients with strabismic amblyopia treated by full-time atropine (n = 38), intermittent atropine (n = 73), or optical (n = 52) penalization participated. METHODS Logarithm of the minimum angle of resolution (logMAR) visual acuity, and binocularity index were determined. RESULTS All three forms of penalization produced statistically significant mean reduction in amblyopia (1.7-2.7 logMAR lines) and mean improvement in binocularity by the end-of-treatment or long-term follow-up visit or both, with minimal mean loss after discontinuation or slight mean improvement on these measures at long-term mean follow-up of 1.9 to 4 years across groups. Few patients achieved high-grade stereoacuity. Compliance was high. Comparable efficacy was found for all three treatment groups after controlling for age, depth of amblyopia, and binocularity at the initial visit. Initial-visit amblyopia depth was strongly and significantly associated with amblyopia depth at both post-treatment visits. Pretreatment and post-treatment binocularity showed a similar strong relationship. Surprisingly, however, there was no consistent or significant association found between depth of amblyopia and binocularity in any visit combination. Post-treatment measures of these two variables also were not associated with initial-visit age or refractive error at any clinically significant level. Mean treatment duration was 1.1 to 2.9 years and was not found to be associated with visual outcome. Amblyopia reversal was found in one (full-time atropine) case at a clinically important level. CONCLUSIONS The authors confirmed previous reports of penalization's efficacy as a primary treatment of moderate amblyopia (20/100 or better acuity) and, in some cases, relatively severe amblyopia (>20/100) and also confirmed its ability to significantly improve mean binocularity. Amblyopia and binocularity appear to respond to treatment independently and, within the postinfancy age range of the sample studied, the responses appear to be independent of initial-visit age. The high acceptability to patients and parents of atropine penalization, and particularly of the intermittent regimen introduced here, suggests the need for prospective-study-based re-evaluation of the relative merits of penalization and occlusion as the standard of care for mild-to-moderate amblyopia.