Twenty-six workers in a hard metal manufacturing plant were monitored by cobalt urinary and ambient air measurements during the first month after summer holidays Cobalt determinations were performed utilizing AAS, with a preliminary chelation and extraction procedure for urinary samples Almost all personal ambient air samples turned out to be under the cobalt dust TLV of 0 1 mg/m 3 When restarting work after the holidays, urinary values did not differ from the control group At the end of the first working week, urinary cobalt had increased four fold, then decreased to the original values on the following Monday before restarting work Thereafter, the weekend was no longer sufficient to reduce the levels to normal urinary cobalt values The values rose to the same level observed before the holidays, and dit not substantially decrease even after the weekend End-shift urinary cobalt values showed a good relationship with present as well as with mean past exposure on the first and the fifth weekday, but the third day did not The correlation was better with present exposure on Monday and with mean past exposure on Friday The observed differences may be explained by the minor influence of recent exposure on present exposure on Monday. The highest values were found on Wednesday We suggest the utilization of end-shift urinary cobalt determination as a measure of the present exposure on Monday, and of mean recent or preceding exposure on Friday.