Topical retinoids are believed to increase inflammatory lesions within the first few weeks of treatment. We evaluated data from several clinical trials for evidence of a signal for retinoid aggravation of inflammatory lesions using a psychometric method and the proportion of participants who demonstrated varying degrees of increased lesion counts. We first determined the validity of a psychometric method based on Stevens’ power law called the visual logarithmic scale (VLS) used to evaluate the perceived changes in inflammatory lesions. There was concurrence between the VLS model and the dermatologists’ visual assessment of a flare in 80.0% (32/40) of participants (P=.0258). A subsequent analysis was performed using data from clinical trials to assess the occurrence of flares using the VLS model or percentage-based definitions (5%, 10%, or 20% increase) following the first week of treatment with various adapalene gel formulations. In this analysis, no evidence of worsening or a flare was seen by either the VLS model or percentage-based definitions. The VLS model is valid for assessing the changes in acne severity. Topical retinoid treatment was not associated with a flare as measured by either the VLS model or the proportion of participants who showed an increase in inflammatory lesions.