A quantitative in situ hybridization technique (quantitative whole-mount in situ hybridization, QISH) for plants is described. It employs direct hybridization of fluorescently labelled gene-specific oligonucleotides in large tissue pieces combined with optical sectioning. It dramatically increases the throughput compared with conventional antibody- and microtome-based in situ mRNA hybridization methods, while simultaneously eliminating artefact-prone preparation steps that prevent reliable quantification in conventional methods. The key feature of this technique is the quantification of gene expression using housekeeping genes (cytosolic GAPDH and 18S RNA) as internal standards. This feature enables a correction of varying cytoplasm/vacuole ratios in different cell types, as well as tissue optical effects and non-specific signals. The quantitative nature of the technique allows for analysis of gene expression in response to different environmental conditions, as well as tissue- and age-dependent differences in gene expression patterns. In addition to testing tissue permeabilization, structural preservation, specificity, linearity and tissue optical effects, we verified the reliability of the technique with three Arabidopsis thaliana genes of known function and distribution. These were the rbcL gene for ribulose 1,5-bisphosphate carboxylase, the developmentally related gene SCARECROW (AtSCR) and PHOT-1, a photoreceptor kinase. As expected, rbcL mRNA was found in all photosynthetic cells, while SCR mRNA was detected mainly in bundle sheath cells and PHOT-1 was found predominantly in epidermal and cortical cells of the apical hook of light-grown seedlings. As an application example, QISH was used to measure transcript abundance for a zinc transporter from the ZIP family of transporters in the Zn/Cd hyperaccumulator model plant, Thlaspi caerulescens, and the related non-accumulator Thlaspi arvense. This showed that QISH can be used to compare differences in mRNA levels between cell types, plant growth conditions and plant species. Messenger RNA for the zinc transporter gene ZNT1 was abundant in photosynthetic cells, but not in the epidermal storage cells where metal hyperaccumulation in T. caerulescens occurs. This indicates that ZNT1 does not directly participate in metal hyperaccumulation within the leaf. Growing T. caerulescens with high zinc levels strongly reduced ZNT1 transcript abundance in the spongy mesophyll cells, but less in the other cell types. In T. arvense, ZNT1 mRNA levels were generally much lower, and were furthermore drastically reduced by growth at increased zinc levels, confirming earlier reports regarding ZNT1 regulation in these two Thlaspi species.