Summary The acute toxicity, pharmacokinetics and hypoxic cytotoxicity of RSU-1069 were investigated using the subcutaneous (sc) rat 9L tumour model. The pharmacokinetics were studied after i.p. injection of RSU-1069 (20mgkg-' or 100mgkg-'). For both doses, the elimination of RSU-1069 followed first-order kinetics in both plasma and unclamped tumours. After 100mgkg-', the peak plasma concentration of RSU-1069 was 40 lgml-'; the elimination ti was 39.3± 11.lmin. After 20mgkg', the peak plasma concentration was 3 pg ml-'; the elimination 4 was 47.8 ± 6.3 min. In unclamped tumours, the peak concentration was 50 jig g -' with an elimination 4 of 36.1 ± 9.6 min for the 100 mg kg -' dose, and 4 tg g-with an elimination 4 of 41.9 ± 6.1 min for the 20 mg kg-' dose. The tumour and plasma elimination half-times were not significantly different (P> 0.2) for the two doses. Clamping the tumour 30 min after administration of 100mg kg-' of RSU-1069 decreased the tumour elimination t to 10.9 ± 1.4 min. After releasing the clamp, RSU-1069 returned rapidly to the unclamped tumour concentration. The unclamped tumour/plasma ratio reached a maximum of 4-6, then decreased to a constant value of about 2 for both doses, indicating that RSU-1069 accumulates in these 9L tumours. RSU-1069 kills hypoxic sc 9L cells more efficiently than oxic sc 9L cells; at a surviving fraction of 0.5, the SER was 4.8. For in vitro 9L cells, the SER was ; 50 when the comparison was between those treated in 2.1% 02 and those treated in <7.5 x 10 3% 02; it was ; 100 when the comparison was between those treated in 21% 02 and those treated in <7.5 x 10-3% 03. Tumours treated with RSU-1069 and clamped for various times exhibited biphasic cell-kill kinetics; at 50mg kg-', little additional cell kill was achieved after 40 min of clamping. Our data also indicate that RSU-1069 is 300-1000 fold more efficient than misonidazole or SR2508 for killing hypoxic sc9L tumour cells in situ.