Much interest has been shown in the use of multi-template reverse transcription-polymerase chain reaction (RT-PCR) as a quantitative instrument for low-abundance mRNAs. A desire to achieve finely-graded quantification of the stress- and hormone-related regulation of one splicing decision in an ion channel gene motivated us to test the reliability of simultaneous amplification of two splice variants with one pair of flanking constitutive primers. Unexpectedly indiscriminate heteroduplexing between the two amplification products, despite a large length difference, and their tight comigration with one homoduplex, mandated a rigorously-denaturing electrophoresis protocol. Conveniently, a new fluorescent dye with high affinity for single-stranded DNA has become available. Though the dye has a good dynamic range, we found that dye and gel saturation compounded by the length difference between products introduced an asymmetrical error into the calculation of relative abundance. Avoiding several pitfalls, dye calibration could be used to correct the error. We also found that differences in the amplification efficiency of the two templates were not constant, but dependent on the initial template ratio, requiring a non-linear correction. Together these improvements gave us very consistent quantitative results, and thus advance our analysis of hormonal mechanisms underlying the regulation of alternative splicing of an ion channel critically involved in stress responses.