Cobalt is a metal that forms several insoluble oxides and a series of water-soluble bivalent salts. Cobalt soaps contain cobalt with various lipophilic carboxylic acids. The most significant occupational exposure is to cobalt metal in the hard metal industry. Following inhalation of cobalt oxides, mucociliary clearance from the tracheobronchial region and dissolution of uncleared particles in pulmonary macrophages is followed by absorption of the solubilised cobalt into the blood. For cobalt (II, III) oxide, removal of particles to the lymph nodes and binding of cobalt to lung tissue may also occur, the latter leading to some retention in the lungs. Little oral absorption of cobalt oxides occurs; for cobalt salts values of 5-30% have been reported. Distribution occurs at highest levels to liver and kidney, and there is some concentration in myocardium relative to skeletal muscle. Following initial excretion secondary to mucociliary clearance, excretion of inhaled cobalt is in the urine. There is little information on the toxicokinetics of cobalt following dermal exposure, although systemic absorption of soluble salts has been reported. Little useful information is available on the effects of single exposure. Studies involving intratracheal instillation indicate only that cobalt metal is considerably more toxic than other components of hard metal, causing an inflammatory response and possibly fibrosis. Cobalt metal, cobalt (II, III) oxide and cobalt naphthenate have low acute oral toxicity in rats, while cobalt (II) oxide and cobalt salts are moderately toxic. No information is available on skin and eye irritation in animals. Cobalt salts cause skin sensitisation in guinea pigs. A number of studies have investigated the effects of repeated exposure. In a satisfactory study, miniature swine inhaled cobalt metal (0.1 or 1.0 mg/m3, particle size 0.4-3.6 um) for 6 hours/day, 5 days/week for 12 weeks. Pulmonary function tests revealed decreased compliance in cobalt-exposed animals. Fibrosis in alveolar septa showed a concentration-dependent increase, and a noeffect level could not be identified. In an adequate study, hamsters inhaled cobalt (Il) oxide (8 mg Co/m3, mass median particle diameter = 045 um) for 7 hours/day, 5 days/week for up to 20 months. Pneumoconiosis characterised by fibrosis of alveolar septa, emphysema, epithelial hyperplasia and metaplasia developed. In a well-conducted study, rats and mice were exposed to cobalt sulphate (0.1, 0.4,1.1, 3.8 or 11.4 mg Co/m3, mass median aerodynamic diameter 0.8-1.1 um) for 6 hours/day, 5 days/week for 13 weeks. In both species effects were seen in the nose, including degeneration of the olfactory epithelium, from 3.8 mg Co/m3. In the larynx, squamous metaplasia was seen from 0.1 mg Co/m3 while, in rats only, inflammatory polyps occupying up to half the laryngeal lumen occurred from 3.8 mg Co/m3. Effects seen in the lungs at 3.8 and 11.4 mg Co/m3 included fibrosis around bronchioles and in alveolar septa, alveolar emphysema, regeneration of bronchiolar epithelium and alveolar epithelial hyperplasia. Intra-alveolar macrophage infiltration was apparent from 0.1 mg Co/m3. A no-observed-adverse effect level could not be identified by this study. In a series of inhalation studies using cobalt chloride (0.4-2 mg Co/m3, mass median aerodynamic diameter 0.5-1 um), rabbits showed hyperplasia of type 11 alveolar cells at 6 weeks, with nodular aggregations seen at 16 weeks. Pulmonary macrophages of cobalt-exposed animals were larger, more numerous and more active. Fibronectin in lung lavage fluid was increased. Several oral studies generally conducted with cobalt sulphate (3-100 mg Co/kg/day for 3 days-20 weeks) have investigated the role of cobalt salts in causing beerdrinkers' cardiomyopathy. No dermal studies are available. Studies with soluble cobalt salts, the most readily bioavailable form of cobalt, provide some evidence of mutagenicity in mammalian cells in vitro and in vivo, although this evidence is not particularly convincing. Cobalt chloride and cobalt naphthenate proved negative in bacterial (Ames) tests for point mutations. Two of the 3 point mutation assays available for cobalt chloride in mammalian cells indicate mutagenic activity, the other being negative, but the brevity of the reporting and the possibility of a confounding influence of cytotoxicity make the positive results unconvincing. In a totally inadequate study, cobalt nitrate did not cause chromosomal aberrations in human cell cultures, in the only available in vitro assay for clastogenicity. A recent in vivo study in mice seems to provide clear evidence of clastogenicity for cobalt chloride. However, the results are unusual in that the large increases seen in chromosomal aberrations were not only dose-related but also occurred at all fixation times used, including one as early as 6 hours after dosing. A briefly-reported micronucleus test in mice was negative. A lifetime inhalation study of fair quality of cobalt (II) oxide in hamsters did not produce any evidence for carcinogenicity. In a recent study involving a series of intratracheal instillations in rats over 2 years, however, the oxide caused a low incidence of malignant pulmonary tumours. No such tumours were seen in controls, but because of the route of administration the significance of the findings was uncertain. No other study of carcinogenic potential using a route at all relevant to occupational exposure is available for cobalt metal, cobalt salts or cobalt soaps. Although cobalt metal causes sarcomas in rats at the site of single injections into muscle, the significance of these findings is very doubtful. Similarly, the positive results obtained in rats and mice following parenteral administration of cobalt chloride or cobalt (II) oxide are of very little relevance in relation to occupational exposure. No information is available on effects on the reproductive system and reproduction for cobalt metal, cobalt oxides or cobalt soaps Effects including testicular atrophy and decreased sperm motility were reported in mice (but not rats) in a study in which the animals inhaled cobalt sulphate at 0.1-11.4 mg Co/m3 for 6 hours/day, 5 days/week for 13 weeks. Similarly, cobalt chloride (2372 mg Co/kg/day) given in drinking water to male mice for 12 weeks caused decreased testis weights, and large decreases in the fertility of the mice. Degenerative and necrotic changes were seen in the seminiferous tubules when rats received cobalt chloride (20 mg Co/kg/day) in the diet for 10-14 weeks. In general, no-effect levels cannot be identified for testicular effects in these studies. Two studies involving administration of cobalt chloride (3-25 mg Co/kg/day) by gavage to pregnant rats do not provide any evidence for teratogenicity, even at maternally-toxic doses. Intraperitoneal administration of cobalt salts led to teratogenic effects in mice and hamsters, but the significance of these findings is unclear. Cobalt chloride has occasionally caused short-term gastro-intestinal effects in children following ingestion. Cobalt metal, cobalt salts and cobalt naphthenate have been reported to cause allergic dermatitis in various industries. In the hard metal industry the greatest risk of cobalt sensitisation comes from activities that are traumatic to the hands, and workers with a sensitivity to nickel are predisposed to cobalt sensitivity. Repeated occupational exposure to cobalt-containing atmospheres in various industries has resulted in two forms of lung disease - diffuse interstitial pulmonary fibrosis and asthma. Numerous cases of each condition have been reported. Cobalt (as the metal dust or as solubilised ionic cobalt) is the probable causative agent in these industries, of which hard metal manufacture and use is the most significant. The mechanism(s) by which cobalt produces these lung effects is(are) uncertain. Immunological responses have been implicated in both the diffuse interstitial pulmonary fibrosis and in the asthma, and there is certainly direct evidence that respiratory sensitisation is responsible for at least some of the asthma cases. However, other mechanisms (eg irritancy, cytotoxicity) may also be involved in the production of these diseases. From the literature available it is not possible, with any confidence, to deduce the dose-response relationships and to distinguish "effect" from "no-effect" levels for cobalt in relation to these diseases. Most of the literature describes individual case reports, either with no data on exposure levels or where is doubtful whether the exposure levels cited give an accurate indication of the personal exposures received by the individuals in question. The health and hygiene surveys conducted have been mainly cross-sectional studies, where contemporary exposure levels were measured but previous exposure levels were unknown. Furthermore, in individuals who have been sensitised to cobalt by inhalation, producing an asthma response, reactions may occur at very low exposure levels. Although addition of cobalt sulphate to beer is implicated in the cause of brief epidemics of cardiomyopathy that occurred in the mid-1960s, other factors were involved and in their absence cobalt cardiomyopathy appears to be rare. Two reports of men dying from cardiomyopathy following industrial exposure to cobalt are available. Despite heavy occupational exposure to cobaltcontaining dust in the past, very few cases of cancer have been reported in cobalt-exposed workers. Overall, from the limited information available there is no clear evidence that cobalt metal and/or its compounds is carcinogenic in humans. There is no information on effects on reproduction in humans.