In multicellular organisms such as Caenorhabditis elegans, differences in complex phenotypes such as lifespan correlate with the level of expression of particular engineered reporter genes. In single celled organisms, quantitative understanding of responses to extracellular signals and of cell-to-cell variation in responses has depended on precise measurement of reporter gene expression. Here, we developed microscope-based methods to quantify reporter gene expression in cells of Caenorhabditis elegans with low measurement error. We then quantified expression in strains that carried different configurations of Phsp-16.2-fluorescent-protein reporters, in whole animals, and in all 20 cells of the intestine tissue, which is responsible for most of the fluorescent signal. Some animals bore more recently developed single copy Phsp-16.2 reporters integrated at defined chromosomal sites, others, "classical" multicopy reporter gene arrays integrated at random sites. At the level of whole animals, variation in gene expression was similar: strains with single copy reporters showed the same amount of animal-to-animal variation as strains with multicopy reporters. At the level of cells, in animals with single copy reporters, the pattern of expression in cells within the tissue was highly stereotyped. In animals with multicopy reporters, the cell-specific expression pattern was also stereotyped, but distinct, and somewhat more variable. Our methods are rapid and gentle enough to allow quantification of expression in the same cells of an animal at different times during adult life. They should allow investigators to use changes in reporter expression in single cells in tissues as quantitative phenotypes, and link those to molecular differences. Moreover, by diminishing measurement error, they should make possible dissection of the causes of the remaining, real, variation in expression. Understanding such variation should help reveal its contribution to differences in complex phenotypic outcomes in multicellular organisms.