There is a 2-fold increase in the percentage of gonadotropes bearing LH beta or FSH beta mRNAs or antigens as the cells approach proestrus. The purpose of this study was to identify the source of these cells with dual labeling techniques. The first hypothesis was that they stemmed from small monohormonal gonadotropes (containing only LH or FSH) that were driven to transcribe and translate the other gonadotropin. Alternatively, they may stem from other pituitary cell types. We detected the LH beta or FSH beta mRNAs by in situ hybridization (biotinylated oligonucleotide probes were detected by peroxidase-labeled avidin). Second, immunolabeling protocols localized the pituitary hormones. The percentages of cells with LH beta antigens and FSH beta mRNA increased to 81% of the LH beta antigen-bearing cells by the time of peak expression of FSH beta mRNA. Similarly, FSH beta antigen-bearing cells increased their expression of LH beta mRNA to 40% of such cells by the morning of proestrus. During the peak period of expression (the evening of proestrus), LH antigen-bearing cells had increased their expression of LH beta mRNA to 93%. Furthermore, 81% of the same cells expressed FSH beta mRNA. Thus, at least 80% of cells with LH antigens became bihormonal as the cells approached proestrus. This partially supports the first hypothesis for the origin of the new gonadotropes. However, the dual labeling studies also showed that 47% or 60% of cells with GH antigens expressed LH beta or FSH beta mRNAs, respectively, during peak expression (14% of pituitary cells contained gonadotropin mRNAs and GH antigens). Expression by cells with other antigens was low or absent (< 5% of pituitary cells). Perhaps a subset of somatotropes expresses gonadotropin mRNAs. Alternatively, the labeling could signify the presence of GH bound to GH receptors in gonadotropes. In either case, it appears that GH cells may be somehow linked to gonadotrope function as the cells approach proestrus.