The paper discusses the use of within-array replicate spots in microarray experiments to improve the assessment of differential expression. The authors propose a method to estimate the correlation between replicate spots, which is then used to improve the precision of estimating genewise variances. This method is validated using data from a microarray experiment involving calibration and ratio control spots, showing that it significantly improves the discrimination of differentially expressed genes. The method is implemented in the limma software package for R. The paper also explores the combination of this method with empirical Bayes methods to further enhance the reliability of the estimates.The paper discusses the use of within-array replicate spots in microarray experiments to improve the assessment of differential expression. The authors propose a method to estimate the correlation between replicate spots, which is then used to improve the precision of estimating genewise variances. This method is validated using data from a microarray experiment involving calibration and ratio control spots, showing that it significantly improves the discrimination of differentially expressed genes. The method is implemented in the limma software package for R. The paper also explores the combination of this method with empirical Bayes methods to further enhance the reliability of the estimates.