Genetically identical cells in the same environment can have large cell-to-cell differences in gene expression. A major reason for these differences is that transcription occurs in stochastic “bursts” of activity. Stochastic bursts of expression diversify cell types during development, contribute to chemotherapeutic resistance in cancer cells, and hinder the production of pure cell types in reprogramming experiments. Although bursting is an important property of gene expression in development and disease, the molecular causes of cell to- cell variability in expression remain poorly understood. Both the mean expression levels and cell-to-cell variability of genes are important in development. A key experimental observation is that genes with the same mean expression can have different cell-to-cell variability. Therefore, molecular mechanisms must have evolved to uncouple the regulation of cell-to-cell variability from mean expression levels. How cells can control cell-to-cell variability without influencing mean expression is unknown. We propose to use the “Bursty Model” of expression as a framework for separating the regulation of cell to- cell variability from mean expression levels. The Bursty Model suggests that the mean and variability of expression can be separated through independent control of a gene’s “burst size'' (the number of mRNAs produced per burst) and “burst frequency” (the time between bursts). Genes producing large but infrequent bursts (developmental regulators) would have both high mean levels and high cell-to-cell variability, while genes with smaller more frequent bursts (metabolic enzymes) could reach the same mean levels with much less variability. Therefore, identifying the molecular mechanisms that separate the control of burst size and frequency would be an important step towards explaining the independent control of mean and variability in expression. We propose to determine whether perturbations to different transcription factor (TF) functions have separable effects on burst size and frequency. Successful completion of our aims would help uncouple the regulation of mean levels from cell-to-cell variability. The results would 1) inform whether natural selection can alter cell-to-cell variability independent of mean levels, 2) shed light on regulatory mechanisms that generate high variability during cell fate decisions, and 3) suggest ways to suppress cell-to-cell variability in applications that rely on the ectopic expression of TFs.