Laboratory tests can provide useful information about the presence and effects of HIV-1 in the CNS, but have thus far not yielded definitive diagnostic or prognostic markers of HIV-1-related cognitive and motor complex. The most clinically useful laboratory procedures are MR imaging and CSF examinations. The routine clinical use of MR imaging and CSF examinations, however, is still restricted to providing information for detecting and excluding secondary effects of HIV-1 infection. MR imaging and CT do not appear to be sensitive enough at current resolutions to provide early detection of HIV-1 CNS effects nor to follow disease progression. Several CSF variables are extremely promising as early markers of primary HIV-1 infection of the brain, and may provide preclinical indications for onset of treatment and for evaluation of treatment efficacy. These include CSF quinolinic acid levels, acid dissociated p24 antigen levels, neopterin or beta 2m, intrathecal IgG synthesis rate, and possibly quantitated PCR levels of HIV-1 viral load. Procedures such as nuclear magnetic resonance spectroscopy, SPECT, PET, computerized EEG, EP, and ERPs are all promising candidates for early detection or localization of HIV-1-related brain dysfunction, but at this time all must still be considered primarily research tools. Before any of these procedures can provide reliable diagnostic and prognostic information about primary HIV-1 neurologic disease, currently on-going longitudinal evaluations of large numbers of asymptomatic HIV-1-infected individuals as they progress to neurologically symptomatic disease must be completed. There is currently no laboratory marker in blood or CSF that definitively predicts the risk for HIV-1-associated cognitive/motor complex. HIV-1-associated cognitive/motor complex remains a clinical diagnosis, which is made on the basis of positive neurologic signs and symptoms and abnormal neuropsychological findings after other causes of neurologic disease are excluded. Laboratory measures, such as the electrophysiologic methods and some CSF variables, are likely to remain adjuncts to the diagnosis because, with few exceptions, they provide data that are nonspecific as to etiopathogenesis. Dynamic imaging, electrophysiologic methods, and CSF indices provide presumptive evidence for the presence of HIV-1-associated CNS damage, and with clinical and neuropsychological evidence, could be used to establish a new definition of primary HIV-1-associated CNS disease along the lines used in establishing a diagnosis of multiple sclerosis.(ABSTRACT TRUNCATED AT 400 WORDS)