The aim of the current study was to compare the objective response and survival rates of patients with mRCC treated with IL-2 administered either systemically (SYST, subcutaneously) or via inhalation (INH), using relatively large sample sizes to afford a more meaningful comparison. We used univariate and multivariate analyses to retrospectively evaluate the data from two different databases generated from 277 patients treated with IL-2 during the 1993-1997 period, one developed at the University Hospital Hamburg-Eppendorf, and the other at Chiron-Amsterdam. Patients treated with INH IL-2 tended to have a poorer ECOG performance status than patients receiving SYST IL-2. Of 75 patients receiving INH IL-2, eight (10.7%) achieved an objective response; of 202 patients administered SYST IL-2, 45 (22.2%) achieved an objective response. The median survival time was 13.8 months for patients receiving INH IL-2 and 13.1 months for patients treated with SYST IL-2. One- and two-year survival rates were also comparable for the two treatment modalities (one-year: INH, 55%; SYST, 56%; two-year: INH, 28%; SYST, 26%). There was no significant difference in the likelihood of survival for patients receiving INH IL-2 versus SYST IL-2 (risk ratio = 0.82, P = 0.27). Patients administered INH IL-2 experienced considerably less toxicity and complications than patients administered SYST IL-2. We conclude that INH IL-2 treatment is at least as effective as SYST IL-2 treatment in promoting the survival of patients with mRCC. Given that INH IL-2 treatment of patients with a poorer ECOG performance status elicited a survival rate comparable to that seen with SYST IL-2 treatment of patients with a superior performance status, the potential exists for INH IL-2 treatment to be even more effective for patients having a better performance status. Additionally, INH IL-2 treatment is considerably less toxic and associated with fewer complications than SYST IL-2 treatment, thus providing a therapeutic option for otherwise untreatable patients, offering patients a relatively good quality of life, and requiring fewer co-medications. Nonetheless, selection of an IL-2 treatment modality should be based on several patient-related considerations. Moreover, these two IL-2 treatment modalities need not be mutually exclusive.