Despite widespread public health campaigns and increased knowledge of the harmful effects of drinking during pregnancy, greater than 1% of children are estimated to have prenatal alcohol exposure. Reading-related difficulties are of particular concern in the school-age population. The current study aimed to characterize reading performance in children with heavy prenatal alcohol exposure. Children (6–12y) with histories of heavy prenatal alcohol exposure (n=32) and without (n=40) were administered a two-hour neuropsychological battery, which included multiple measures of reading domains (decoding, fluency, comprehension) and phonological processing (phonological awareness, phonological memory, rapid naming). Caregivers completed assessments of home literacy environment and behavior. Correlation, MANOVA, and regression techniques were conducted to evaluate differences between groups and identify contributing factors for reading performance. Discriminant function and latent class analyses were run to determine whether performance on these measures could aid in differential diagnosis and establish whether distinct subtypes of reading impairment exist. There were no significant differences on demographic characteristics between groups. Alcohol-exposed children performed significantly worse than their peers on all measures, with the exception of rapid naming. In particular, alcohol-exposed children had relative weaknesses in phonological awareness, decoding, and comprehension. They also had significantly higher rates of reading difficulties in all domains. Aspects of phonological processing accounted for significant variance in reading variable across groups. Exposure history accounted for additional variance in decoding and comprehension. No interaction effects were significant. After other factors were added to the models, vocabulary, behavioral concerns, and attitude towards reading were additional significant contributors. Also, exposure history continued to account for significant variance. Outcome variables distinguished between alcohol-exposed children and controls. Distinct subgroups emerged based on severity of impairment. There were no significant differences between performances on academic domains (reading, spelling, math) for either group. Alcohol-exposed children had significant difficulties in all aspects of reading, comparable with their performance in math and spelling. They demonstrated specific weaknesses that suggest potential targets for intervention. Cognitive mechanisms that contribute to reading in both typical and neurodevelopmental disorder populations were also found in this population. Therefore, effective interventions in other populations may be utilized to improve outcomes for alcohol-exposed children.