The use of reverse transcription polymerase chain reaction (RT-PCR) analysis of melanoma-specific transcripts for the identification of circulating melanoma cells has shown very variable results in different studies on melanoma patients. We have therefore developed quantitative methods to study both analytical and biological variations as possible causes of this phenomenon. Pigment-related and S-100 beta transcripts were quantified in 12 different melanoma cell lines and related to the amounts of 5-S-cysteinyldopa, pigment and S-100B protein. A real-time PCR method was used and the results were expressed as absolute number of transcripts per cell. Tyrosinase, tyrosinase-related protein (TRP)-1, TRP-2 and MART-1/Melan-A mRNA varied from undetectable (< 10(-4) transcripts/cell) to 10(3) transcripts/cell, i.e. by a factor > 10(7) in the different cell lines. S-100 beta mRNA varied from 2.8 to 165 transcripts/cell, i.e. by a factor of 60. Tyrosinase, TRP-1 and TRP-2 mRNA correlated significantly with the amount of 5-S-cysteinyldopa, an intermediate pigment metabolite (P < 0.001, P < 0.001 and P < 0.01, respectively). The amount of S-100 beta mRNA correlated significantly with the amount of S-100B protein (P < 0.001). No cross-correlations were seen between the pigment-related and S-100-related analytes. We conclude that one reason behind the negative results of RT-PCR measurement of pigment-related mRNA may be that these transcripts are not always expressed in the particular cells present in the patient's blood. Furthermore, variation in the expression of the order of 10(7) must have great impact on the diagnostic sensitivity. Measurement of S-100 beta mRNA would be more sensitive, but the use of this transcript is hampered by its presence in the blood cells.