PURPOSE: To determine whether the fluxes through hairless mouse skin for three homologous series of prodrugs of 5-fluorouracil (5-FU, 1) and 6-mercaptopurine (6-MP, 2) from saturated aqueous suspensions show dependencies on aqueous (SAQ) and isopropyl myristate (SIPM) solubilities similar to those shown by the identical compounds delivered from IPM. METHODS: Flux through hairless mouse skin from water (JMAQ) and solubility data were measured for a homologous series of six 3-alkylcarbonyloxymethyl (ACOM) prodrugs of 5-FU (3-ACOM-5-FU), and five 6-ACOM-6-MP prodrugs, then combined with literature data for five bis-6,9-ACOM-6-MP prodrugs to give a data base. Multiple linear regression using SPSS 7.5 was performed on log SIPM, log SAQ, molecular weight and log JMAQ data to determine the best fit coefficients to the transformed Potts-Guy equation: log JMAQ = x + y log SIPM + (1 - y) log SAQ + z MW. Permeability coefficients (PMAQ) were calculated from JMAQ/SAQ. RESULTS: The best fit coefficients for the flux from AQ(JMAQ) were x = -1.497, y = 0.660 and z = -0.00469 (r2 = 0.765) with an average error of prediction equal to 0.193 log units. The best fit coefficients for the flux from IPM (JMIPM) were x = -0.557, y = 0.536 and z = -0.00261 (r2 = 0.941) with an average error of prediction equal to 0.109 log units. For all three series, log PMAQ increased whereas log PMIPM decreased with increasing alkyl chain lengths in the promoiety and with decreasing solubility parameter values. CONCLUSIONS: The transformed Potts-Guy equation can be used to predict JMAQ but with less certainty than JMIPM. SIPM and SAQ have consistently been shown to have a positive influence on JMIPM, and now on JMAQ, with a balance between the two solubilities being obviously important. The previous observation that log PMAQ increased with lipophilicity is an artifact of normalizing JMAQ by SAQ.