In a previous paper(1) the background, organisation and early results of a collaborative project on the safety of optical measurements in hazardous industrial environments involving five European laboratories in France, Germany and the UK were described. It is necessary that safe operating power levels should be established for the peace of mind of both manufacturer and user of optical sensors, and paradoxically, to enable an intensity of radiation to be employed which does not disadvantage light as a measurement medium and thereby act as a technological constraint. In general it can be assumed that the more light that goes into an optical measurement system the more utilitarian will be the result. Technically, determining safe operating radiation levels is a difficult task because of the many variables involved. We have concentrated on a single mechanism: thermal ignition of an atmosphere surrounding a body which is heated by the incident radiation. Influential factors are the nature of the absorber (size, shape, form, chemical reactivity), the nature of the environment (air, dust, level of flammable gas present) and the nature of the incident radiation (wavelength, coherence, beam geometry and temporal variation). Because of the breadth of this problem experimental investigations are being supplemented by theoretical calculations in the hope of realising a predictive capability which can anticipate hazards for situations for which experimental results are not available. It is the objective of this paper to present an interim statement on the results obtained up to the half-way point in this 3-year European project. This paper draws heavily on the work of one of the collaborating laboratories, a fuller account of which will shortly appear in the scientific literature2. More importantly it builds bridges to the other relevant results which have been published over the last several years. Of particular relevance are the publications from the Australian 'school'.