Search for beryllium (Be) in tissues or urine in suspected beryllium disease is often disappointing due to inferior sensitivity of the methods employed. We evaluated the clinical use of laser microprobe mass spectrometry (LAMMS) for measurement of Be and detected the metal to a minimum concentration of 1 microM. We then investigated the biological relevance of this concentration. We looked at the alveolar macrophages in 7 patients subsequent to an incubation period of 24 h as well as peripheral blood mononuclear cells (PBMNC) and various cell lines with and without addition of beryllium sulfate (BeSO4). We also investigated skin biopsies of two patients 28 days after intracutaneous injections of BeSO4 (beryllium skin testing) and alveolar macrophages of A/J mice at various time intervals after a beryllium sensitisation protocol and a single intratracheal injection of BeSO4 (maximum interval: 15 weeks). Be was not defectable in native patient alveolar macrophages (AM), but in 6 of 7 cases after coincubation with BeSO4. There was no significant Be signal in any analysed PBMNC sample or cell line--even after coincubation with Be--or in the skin probes. Murine AM, however, had incorporated significant amounts of Be, which were detectable until the end of the experiments 15 weeks later. We conclude that concentrations of Be in acute disease (here, inoculation of Be in mice) exceed 1 microM and are thus detectable by LAMMS. On the other hand, concentrations in chronic processes (in this cosar, skin testing) are below the detection limits of LAMMS. Further results suggest compartmentalisation of the immune processes induced by Be because the alveolar macrophages were able to incorporate Be while PBMNC were not and because AM--at least in the animal experiments--seem to store Be intracellularly.