The purpose of this study is to investigate the theoretical structure of the Survey of Reading Strategies (SORS) developed for assessing ESL students' use of metacognitive reading strategies. The data for this study were collected from 1,431 Chinese college students whose second language (L2) was English. Both explanatory factor analysis (EFA) and confirmatory factor analysis (CFA) were conducted to validate the SORS. The EFA failed to recover the three hypothesized subscales but supported a unidimensional construct. The CFA analysis suggested that all alternative factor models fitted the sample data poorly, but comparative fit indices indicated the construct of the SORS was unidimensional. Therefore, the factor structure of the SORS needs to be further explored. Suggestions and recommendations for the use of the survey were provided.