Several tumour sites have now demonstrated objective responses to alpha interferons in a diversity of doses and schedules. Since effectiveness should be enhanced with the identification of an optimal dose strategy, we undertook a prospectively randomized study to compare an intermittent high dose escalating strategy (HDS) vs. a fixed low dose treatment in relation to clinical outcome and laboratory correlates of immune function. HDS patients received interferon alpha-N1 (lymphoblastoid interferon) 5M units m-2 by continuous i.v. infusion over 24 h, escalating by 5 M units m-2 day-1 as tolerated over 10 days, and repeated every 28 days. The low dose strategy (LDS) consisted of a fixed dose of 2 M units m-2 by intramuscular injection daily for 28 days, then daily for 7 days every other week. There were 53 evaluable patients. In keeping with earlier preliminary results there was evidence of improved immune function for HDS patients. They demonstrated a significant increase in the number of CD2+ (sheep red blood cell binding) cells and CD4+ (helper-inducer/suppressor-inducer) cells along with enhanced activity of natural killer cell, and mixed leukocyte culture activity. In addition to improved immune function, HDS patients survived longer than LDS (P = 0.04). Analysis of survival in relation to response suggested that monitoring of minor responses may be of interest for biological agents such as interferon.