The cold capture assay as described by Brezinsky et al. [Brezinsky, S.C.G., Chiang, G.G., Szilvasi, A., Mohan, S., Shapiro, R.I., MacLean, A., Sisk, W., Thill, G., 2003. A simple method for enriching populations of transfected CHO cells for cells of higher specific productivity. J. Immunol. Methods 277, 141-155] stands out as the most simple of single cell secretion assays which can be used to sort for high productivity in recombinant cell lines. At low temperatures the process of protein release from transport vesicles is assumed to be delayed as both vesicle fusion and product release is slowed, so that secreted proteins can be stained on the cell surface using a fluorescent antibody. Typically, the fluorescent signal obtained correlates to the cell specific production rate of the analysed cell. In the present study we compared staining of human antibody producing CHO cells performed at different temperatures and we observed the fluorescent signal over 24h. We found that the staining temperature did not influence signal intensity. The fluorescent signal was stable for 24h at 4 degrees C, decreased to 80% at room temperature (21 degrees C), while it decreased significantly already after 2h at 37 degrees C. Initially, the fluorescent signal was observed on the cell surface, however, at later stages it was found in compartments in the cytoplasm. Finally we compared differences in signal stability depending on whether the antibody used for staining bound to the light or heavy chain of the product and on whether the fluorescent label was a relatively stable protein (phycoerythrin) or a pH-dependent small molecule (FITC). Our results indicate that the secreted product is trapped by the staining antibody on the cell surface at all temperatures. Subsequently these aggregates are endocytosed by the cells, a process which is slowed down at low temperatures.