The UK reading literacy data from the OECD PISA 2000 study are analysed using multilevel methods. The main analyses focus on the composite reading literacy scale resulting from the study and the explanatory variables used are 16 of the scales derived from answers given to the pupil questionnaire administered as part of the survey, together with pupil gender, and a created variable giving a measure of parental education. A variable distinguishing state and private schools was also entered at the school level. The principal significant school-level variables were those related to socioeconomic status. The variable differentiating between state and private schools was not significant. All pupil-level variables were significant. Gender differences persisted, with girls performing better than boys. A secondary series of analyses examined the three reading subscales of "Reading for information", "Interpreting texts", and "Reflecting and evaluating". The pattern of findings was substantially the same as for the composite scale, though family educational support entered as an additional significant school-level factor for the final 2 of these analyses, and the effect of gender showed interesting variations. (Contains 9 tables.)