Matrix eQTL is a fast eQTL analysis tool that uses large matrix operations for efficient computation. It is significantly faster than existing eQTL analysis tools, performing 2–3 orders of magnitude faster on large datasets while finding the same eQTLs. The tool supports additive linear and ANOVA models with covariates, including models with correlated and heteroskedastic errors. It addresses multiple testing by calculating the false discovery rate (FDR) separately for cis- and trans-eQTLs. Matrix eQTL is available in Matlab and R, and can handle large datasets with up to 10 billion tests. It is particularly efficient in handling datasets with millions of SNPs and tens of thousands of transcripts. The tool's performance is enhanced by using matrix operations and efficient programming languages. Matrix eQTL can analyze a dataset with 573,337 SNPs and 22,011 transcripts in under 12 seconds. It is the fastest tool among tested eQTL analysis tools, with performance remaining stable even when covariates are added to the model. Matrix eQTL is also efficient in handling data loading and analysis, with loading times for the full CF dataset ranging from 3.3 to 18.4 minutes. The tool's performance is influenced by the choice of model and other parameters, with the simple linear regression model taking about 14.6 minutes to analyze the CF dataset. The algorithm for the simple linear regression model involves calculating the correlation between gene expression and genotype, which is done using large matrix operations. The tool also supports Q–Q plots and histograms of all p-values, and can calculate FDR for gene-SNP pairs that pass a user-defined significance threshold. Matrix eQTL is efficient in handling large datasets and is suitable for various eQTL analysis tasks, including the analysis of RNA-seq data and the identification of gene-gene interactions. Future versions of Matrix eQTL may include GPU-based calculations and more complex models, such as multi-SNP models and generalized linear models. The tool is supported by various funding sources and has no conflicts of interest.Matrix eQTL is a fast eQTL analysis tool that uses large matrix operations for efficient computation. It is significantly faster than existing eQTL analysis tools, performing 2–3 orders of magnitude faster on large datasets while finding the same eQTLs. The tool supports additive linear and ANOVA models with covariates, including models with correlated and heteroskedastic errors. It addresses multiple testing by calculating the false discovery rate (FDR) separately for cis- and trans-eQTLs. Matrix eQTL is available in Matlab and R, and can handle large datasets with up to 10 billion tests. It is particularly efficient in handling datasets with millions of SNPs and tens of thousands of transcripts. The tool's performance is enhanced by using matrix operations and efficient programming languages. Matrix eQTL can analyze a dataset with 573,337 SNPs and 22,011 transcripts in under 12 seconds. It is the fastest tool among tested eQTL analysis tools, with performance remaining stable even when covariates are added to the model. Matrix eQTL is also efficient in handling data loading and analysis, with loading times for the full CF dataset ranging from 3.3 to 18.4 minutes. The tool's performance is influenced by the choice of model and other parameters, with the simple linear regression model taking about 14.6 minutes to analyze the CF dataset. The algorithm for the simple linear regression model involves calculating the correlation between gene expression and genotype, which is done using large matrix operations. The tool also supports Q–Q plots and histograms of all p-values, and can calculate FDR for gene-SNP pairs that pass a user-defined significance threshold. Matrix eQTL is efficient in handling large datasets and is suitable for various eQTL analysis tasks, including the analysis of RNA-seq data and the identification of gene-gene interactions. Future versions of Matrix eQTL may include GPU-based calculations and more complex models, such as multi-SNP models and generalized linear models. The tool is supported by various funding sources and has no conflicts of interest.