Summary The cellular distribution of 4-borono-2-["8F]fluoro-L-phenylalanine (['8F]FBPA, an analog of pboronophenylaline), a potential agent for boron neutron capture therapy (BNCT), and [6-3H]thymidine ([3H]Thd, a DNA precursor) in murine two B16 melanoma sublines and FM3A mammary carcinoma was studied in vivo using double-tracer microautoradiography. Tumour volume, tumour age, cell density in the tissues and the proportion of S phase cells in the cell cycle were the same in the three tumour models. Volume doubling time, which represents tumour growth rate, was fastest in B16F10, followed by B16F1 (P <0.05), the slowest being in FM3A (P<0.001). p-Boronophenylalanine (BPA) has been studied as a potential agent for boron neutron capture therapy (BNCT) for melanomas (Mishima et al., 1989a,b). The success of this treatment is dependent on the highly selective localisation and concentration of boron-10 in tumours vs normal tissues. Biodistribution studies and radiation dosimetry based on pharmacokinetics have indicated that BPA acts as a biochemically targeted boron carrier for melanomas in vivo (Barth et al., 1990;Coderre et al., 1988). Double-labelled neutron capture radiograms of '°B-L-BPA and [3H]Thd autoradiograms from the same whole-body section showed that high concentration of boron in the tumour corresponded closely with areas of rapid cell division (Coderre et al., 1987). However, the relationship of BPA accumulation to DNA synthesis and pigmentation is still unclear (Tsuji et al., 1983;Ichihashi et al., 1982;Ishiwata et al., 1992a).In recent studies carried out to accurately determine the in vivo concentration of the compound, '8F labelled borono-FBPA) has been synthesised as a positron emitting tracer which can be quantified non-invasively in vivo by positron emission tomography (PET) (Ishiwata et al., 1991a,b; 1992a,b). Chemically determined '0B-BPA concentration in the Greene's melanomas was comparable to the value estimated with ['8F]FBPA (Yamada et al., 1990).Under a safety light, frozen 5-.m sections were directly mounted on slides coated with NTB2 emulsion. After 4 h of exposure under dry ice cold, the sections were fixed with 5% acetic acid; the autoradiograms were developed in Konidol-X (Konica, Japan), fixed in Kodak general purpose fixer (Kodak, Japan), washed, and dried. Three days after the first ARG for the complete decay of '8F (t1/2= 109.8 min), the second ARG was processed with ET2F stripping film. Under a safety light, the slides with sections and the first autoradiogram were covered with ET2F stripping film; the film-