Occupational Exposure Limits (OELs) are used as an important regulatory instrument to protect workers’ health from adverse effects of chemical exposures. The main objective of this thesis is to study risk assessment practices in the setting of OEL in order to produce knowledge that will help improve the consistency and transparency of OELs. For the purpose of paper I a database of OELs for a total of 1341 substances was compiled. Of these, only 25 substances have OELs from all 18 included organisations while more than one third of the substances are only regulated by one organisation alone. The average level of OELs differs substantially between organisations; the US OSHA exposure limits are (on average) nearly 40 % higher than those of Poland. In paper II six EU member states’ OELs are compared to the European Commission’s OELs. Also within Europe there is a large difference concerning the average level of OELs (35%). The average level of lists tends to decrease over time, although there are exceptions to this. There are also indications that the exposure limits of EU member states are converging towards the European Commission’s OELs. The work presented in paper III identifies steps in the risk assessment that could account for the large differences in OELs for 14 different substances. Differences in the identification of the critical effect could explain the different level of the OELs for half of the substances. But the age of the data review could not account for all the differences in data selection, only one fifth of the documents referred to all available key studies. Also the evaluation of the key studies varied significantly. The aim of paper IV was to investigate how the Scientific Committee on Occupational Exposure Limits (SCOEL) of the European Commission uses assessment factors when proposing health-based indicative OELs. For only one third of the investigated OELs were explicit assessment factors given. On average the safety margin of the recommendations was 2.1 higher when an explicit assessment factor had been used. It is recommended that the SCOEL develop and adhere to a more articulate framework on the use of assessment factors. Paper V focuses on the Derived No-Effect Levels (DNELs) which are to be calculated under the new European Union REACH legislation. It is a comparison of the safety margins of 88 SCOEL recommendations with those of the corresponding worker-DNELs, derived according to the default approach as described in the REACH guidance document. Overall, the REACH safety margins were approximately six times higher than those derived from the SCOEL documentations but varied widely with REACH/SCOEL safety margin ratios ranging by two orders of magnitude, from 0.3 to 58. / QC 20110215