The steady-state level of the neuromodulin transcript in the neuron-like N1E-115 cell line was measured with a method combining reverse transcription and the polymerase chain reaction (RT/PCR). Total RNA was isolated from N1E-115 cells and treated with DNAse to remove residual DNA; cDNA was synthesized from this RNA by priming with random hexamers. For PCR amplification, primers for neuromodulin were designed for regions of the coding sequence that were identical in mouse, rat, and human. In one approach (the 'ratio method'), variations in RNA yield and cDNA synthesis efficiency were controlled for by amplifying a reference (housekeeping) gene (glyceraldehyde phosphate dehydrogenase; GAPDH). To control for inter-experimental variations in PCR amplification efficiencies the data were analyzed on semi-logarithmic plots, with which the relative levels of the starting templates could be determined by extrapolating the plots to cycle number zero (0). In another approach with RT/PCR (the 'spiking method'), the absolute level of N1E-115 neuromodulin cDNA was assessed by adding known amounts of cloned human neuromodulin template to the RT/PCR assay of N1E-115 nucleic acid and comparing the increased yield of product across cycles. When the spike was added at either the cDNA level (in the form of double-stranded DNA) or at the total RNA level (as sense RNA), the levels of N1E-115 calculated were virtually the same: 509 fg and 495 fg of coding region per ug total RNA in confluent N1E-115 cells, respectively. Treatment of N1E-115 cells with 2% dimethylsulfoxide for three days elevated neuromodulin mRNA levels 5.6-fold. Conversely, treatment of N1E-115 cells with 100 nM phorbol myristate acetate for 24 h decreased the level of neuromodulin mRNA by 70%. Under carefully controlled conditions and within certain limits of precision, the RT/PCR method appears to be suitable for assessing the level of low abundance mRNA under various pharmacologically-induced conditions.