This dataset contains baseline Early Grade Reading Assessment (EGRA) data conducted under the USAID/Zambia Education Data activity between November – December 2018. Over 15,000 Grade 2 learners were assessed in one of the seven Government of the Republic of Zambia (GRZ) languages of Instruction (LoI) (Chitonga, Cinyanja, Icibemba, Kiikaonde, Lunda, Luvale or Silozi) as well as in English. The EGRA was conducted in five target provinces (Eastern, Muchinga, North-Western, Southern and Western Provinces). The purpose of the 2018 baseline EGRA is to establish a baseline level from which changes in Grade 2 learners’ performance in the core reading skills can be tracked over time. Each assessment contained seven subtasks, which included; (1) listening comprehension in both the LoI and in English; (2) letter sound identification in the LoI; (3) syllable sound identification in the LoI; (4) non-word reading in the LoI; (5) oral reading fluency in the LoI; (6) reading comprehension in the LoI and; (7) English vocabulary. In addition, assessors also administered a Snapshot of School Management and Effectiveness (SSME), which included head teacher, teacher, and learner questionnaires, along with a school inventory, to establish school characteristics and learner demographics in the sampled schools. The 2018 Baseline EGRA used a stratified sampling methodology to randomly select a representative sample of 816 schools from the five target provinces. Of the 816 schools, 630 were Government of the Republic of Zambia (GRZ)-run primary schools and 186 were community-run schools.