Most RT PCRs are short because Reverse Transcriptase tends to fall off due to secondary structure unless special conditions are used. Tth polymerase itself has some reverse transcriptase activity and is used in some Roche Assays (in the presence of Manganese) for this purpose, but, again, the reverse transcripts are short. Then, you will want to span some Exons, so you need to map these against a RefSeq of your choosing. This is because most current RNA preps use total RNA and therefore hnRNA from the nucleus (unprocessed) would confound your result if you did not specifically design primers spanning an intron (so that you are only measuring spliced mRNA). Finally, we use Primer Express in our lab because most of the rtPCRs are turned into TaqMan (5'nuclease assay) qPCRs for quantitative purposes. Primer Express is specifically designed to help calculate appropriate TMs (melting temperatures) for minor groove binding (MGB), non-fluorescent quencher (NFQ) exact-match probes. Good luck in designing your assay. About 100 - 150 bp spanning one intron should do it.
try FastPCR nd good software with multiple anaysis of restriction site avail, chances of primer dimer and datail of in silico PCR so u would have better idea before amplification about result.....
Most of the time we also use Primer express or primer-blast but lately, I found qPrimerDepot (http://primerdepot.nci.nih.gov/) very useful and efficient.
You can design primers using any of the above suggested software. One important point is always check for the self -complimentarity of the primers while designing.
Primer 3 (an online tool) give 5 suggestion of specific pair primers. In this case (design primer for RT-PCR), you must use a specific primer, or oligo dT (if your target has poly-A at end of mRNA).
Try GeneWarrior (http://genewarrior.com) for designing primers, it's based on Primer3 but has an interactive user interface and is quite straight forward.