The influence of major surgery on HIV disease progression and decline in CD4+ cell count was evaluated in 23 seropositive haemophilia patients. 24 HIV-infected patients served as non-operated controls. In addition, 32 age-matched seronegative subjects were included. The follow-up time was up to 5 years. During the course of the study, eight of the operated (35%) and 11 of the non-operated (48%) subjects developed HIV-related symptoms (P=0.267). The relative risk for developing HIV-related symptoms after surgery was 0.60 (95% CI 0.25; 1.48). A significant decline in CD4+ cell counts was observed in both the surgery (4.0 x 10(6)/l/month, 95% CI 2.0; 6.0 x 10(6), P=0.001) and the non-surgery (4.0 x 10(6)/l/month, 95% CI 2.0; 6.0 x 10(6), P=0.004) seropositive subgroup, but no difference between the two subgroups was seen (P=0.793). HIV (6.0 x 10(6)/l/month, 95% CI 2.1; 9.9 x 10(6), P=0.0005) but not surgery (-1.0 x 10(6)/l/ month, 95% CI -3.0; 0.96 x 10(6), P=0.647) was an independent predictor for the decline in CD34+ cell count. No interaction effect was observed between HIV infection and surgery (P=0.361). The annual amount of factor concentrate used for regular replacement therapy did not influence the decline in CD4+ cell count (P=0.492). We conclude that major surgery may be considered in symptom-free HIV-seropositive haemophilia patients, with CD4+ cell counts > or = 0.20 x 10(9)/l under similar premises as for seronegative subjects.