Census is an algorithm that converts relative RNA-Seq expression levels into relative transcript counts without the need for experimental spike-in controls. It improves the accuracy of single-cell gene expression analysis by enabling more reliable differential expression analysis and revealing changes in gene expression, splicing patterns, and allelic imbalances. Census counts can be analyzed using regression techniques to detect developmentally regulated genes and identify cell fate-dependent gene expression. The algorithm is implemented in Monocle 2, an open-source single-cell analysis toolkit. Census works by estimating relative transcript counts in spike-in-free experiments, which are then used for differential analysis. It demonstrates improved accuracy compared to normalized read counts and TPM, as it better fits the negative binomial distribution underlying RNA-Seq analysis. Census counts enable robust single-cell analysis at multiple levels of gene regulation, including splicing and allelic balance. Census was tested on several datasets, including those from developmental and disease contexts, and showed high concordance with spike-in derived estimates. It also improved the detection of genes regulated during cell fate decisions, such as those involved in the specification of alveolar epithelial cells. Additionally, Census enabled the identification of genes regulated during the branching of dendritic cell trajectories in response to interferon signaling. Census counts also enabled the analysis of differential splicing and allelic balance in single-cell RNA-Seq experiments. It demonstrated the ability to detect changes in splicing patterns in differentiating myoblasts and to identify genes with monoallelic expression in pre-implantation embryos. The algorithm was shown to be more accurate than normalized read counts in detecting gene regulation at the allele and isoform level. Census provides a new method for analyzing single-cell RNA-Seq data, enabling more accurate and robust differential expression analysis. It is implemented in Monocle 2 and is available for use in single-cell analysis. The algorithm has been validated on multiple datasets and has shown improved performance compared to traditional methods. It is a valuable tool for uncovering new mechanisms of gene regulation in development and disease.Census is an algorithm that converts relative RNA-Seq expression levels into relative transcript counts without the need for experimental spike-in controls. It improves the accuracy of single-cell gene expression analysis by enabling more reliable differential expression analysis and revealing changes in gene expression, splicing patterns, and allelic imbalances. Census counts can be analyzed using regression techniques to detect developmentally regulated genes and identify cell fate-dependent gene expression. The algorithm is implemented in Monocle 2, an open-source single-cell analysis toolkit. Census works by estimating relative transcript counts in spike-in-free experiments, which are then used for differential analysis. It demonstrates improved accuracy compared to normalized read counts and TPM, as it better fits the negative binomial distribution underlying RNA-Seq analysis. Census counts enable robust single-cell analysis at multiple levels of gene regulation, including splicing and allelic balance. Census was tested on several datasets, including those from developmental and disease contexts, and showed high concordance with spike-in derived estimates. It also improved the detection of genes regulated during cell fate decisions, such as those involved in the specification of alveolar epithelial cells. Additionally, Census enabled the identification of genes regulated during the branching of dendritic cell trajectories in response to interferon signaling. Census counts also enabled the analysis of differential splicing and allelic balance in single-cell RNA-Seq experiments. It demonstrated the ability to detect changes in splicing patterns in differentiating myoblasts and to identify genes with monoallelic expression in pre-implantation embryos. The algorithm was shown to be more accurate than normalized read counts in detecting gene regulation at the allele and isoform level. Census provides a new method for analyzing single-cell RNA-Seq data, enabling more accurate and robust differential expression analysis. It is implemented in Monocle 2 and is available for use in single-cell analysis. The algorithm has been validated on multiple datasets and has shown improved performance compared to traditional methods. It is a valuable tool for uncovering new mechanisms of gene regulation in development and disease.