The antineoplastic and toxic effects of mitomycin (MC) administered as a single pulse injection were compared with the effects of MC given as infusions over 2, 5, or 24 hours and with the effects of the drug administered in fractionated regimens incorporating 2-20 injections. The antineoplastic effects of MC were assessed using EMT6 mouse mammary tumors implanted in BALB/c mice; tumor response was assessed using both cell survival and tumor growth assays. The host toxicity of MC was assessed by measuring the survival of bone marrow stem cells (CFU-S) and by determining LD50/30's (which reflect primarily marrow injury). The survival curve for cells from tumors treated with varying doses of MC was exponential, with no shoulder which might suggest accumulation of sublethal lesions. No evidence was found that the tumor cells repaired MC damage when left in situ for up to 24 hours after treatment (repair of potentially lethal damage) or that the tumor repaired sublethal damage during times of up to 24 hours between drug treatments or during drug infusions. The responses of tumors treated with single injections of MC, with infusions over 2, 5, and 24 hours, with two injections separated by 5 or 24 hours, and with five daily injections were all similar. The survival curve for CFU-S from mice treated with varying doses of MC was exponential, with no shoulder which might suggest accumulation of sublethal lesions. LD50/30's were the same whether MC was administered as single injections, as infusions over 2, 5 or 24 hours, or as five daily fractions. Longer fractionated regimens, in which mice received daily injections, 5 days per week, for up to 4 weeks, produced less host toxicity but also less antineoplastic effect; the therapeutic ratio for the most protracted regimen tested was lower than that for single injections. The therapeutic ratios achieved with infusions and fractionated regimens were never greater than that obtained with a large single dose, confirming clinical impressions that infrequent, large doses of MC are an optimal method of treatment with this drug.