The introduction of protease inhibitors (PIs) gave a dramatic drop in AIDS-related opportunistic events, mainly due to induced immune reconstitution. Discontinuation of prophylaxis against Pneumocystis carinii is considered safe when CD4 > 200 cells/mm(3). Ideally, we should have specific functional tests for HIV-1-related decisions. We examined viro-immunological profiles, clinical outcome and lymphocyte proliferation (LP) to P. carinii and other antigens in 108 subjects: 28 AIDS presenters with P. carinii pneumonia (PCP) (CD4 < 200 cells/mm(3)), 22 untreated asymptomatic HIV-1-infected patients (CD4 > 200 cells/mm(3)), 44 HIV-1-infected patients immune-reconstituted on antiretroviral regimens and 14 HIV-1-uninfected healthy controls. As regards viral load, there was no significant difference in therapy duration, nadir, or actual CD4, CD8, natural killer or B cell counts in immune-reconstituted patients receiving protease inhibitor (PI)-based versus those receiving PI-sparing antiretroviral regimens. Among subjects showing abnormally low P. carinii-specific LP, three patients receiving a non-nucleoside reverse transcriptase inhibitor (nNRTI) developed PCP despite having CD4 > 250 cells/mm(3). P. carinii-specific LP could be considered for doubtful situations, i.e. for a safer clinical decision of discontinuing or restarting prophylaxis in patients with a low CD4 nadir or experiencing a sudden CD4 decrease under highly active antiretroviral therapy (HAART). HIV-1 PIs, having in vitro aspecific effects against Pneumocystis, could play a clinically significant anti-opportunistic role, thus offering a further benefit in heavily immunosuppressed patients during early stages of antiretroviral therapy.