PURPOSE: The effectiveness of school eye health programmes relies on many factors, including compliance with spectacle wear. The objectives of this study were to determine spectacle compliance in a school vision screening pilot programme in Botswana, and investigate factors predictive of compliance. METHODS: The study was an observational, cross-sectional follow-up of a pilot school screening programme. Unannounced compliance checks were completed after 3-4 months in a convenience sample of 19 schools. Sex, age, school level, visual acuity, and refractive error were analysed using logistic regression to investigate factors predictive of compliance. FINDINGS: Compliance data were recorded for 193/286 (67.5%) children; 62.2% were female and the median age was 15 years (interquartile range 12-17 years). 60.1% of the sample were compliant with spectacle wear. Girls were more likely to be compliant than boys (adjusted odds ratio (aOR) = 2.32, 95% confidence interval (CI) 1.03-5.27). Children at primary and junior secondary school were more likely to be complaint than senior secondary school children (aOR = 16.96, 95% CI 5.60-51.39; and aOR = 3.39, 95% CI 1.39-8.22, respectively). Children with binocular uncorrected visual acuity (UCVA) of 6/7.5 to 6/12 were 2.76 (95% CI1.05-7.23) times more likely to be compliant than children with binocular UCVA of 6/6. CONCLUSION: Compliance was higher in Botswana than previous African studies; however, improvement in this area would increase the effectiveness of the programme. Further investigation into barriers to spectacle wear affecting boys and older children is warranted. A prescribing protocol to avoid low prescriptions - especially where binocular UCVA is 6/6 - is desirable.