Visibility is a driving factor in the operation of a tunnel in normal operation as well as in incident situations. Hence, there is a need for appropriate data to determine the visibility influencing parameters. The paper deals with two topics, but both concern visibility. One focus is the design criteria for ventilation in normal mode. While in former years visibility was determined by the emission of combustion related particles (mainly soot), while today - due to the improved emission standards of vehicles - the particles in a tunnel result from the vehicle exhaust (combustion) and from so called non-exhaust emissions (e.g. road abrasion, tire wear, resuspension of road dust, etc.). In an urban area the proportions of exhaust and non-exhaust participate matter (PM) are currently almost 1:1 while in a tunnel the PM count can be dominated by the non-exhaust particulates (depending on the traffic composition). The main problem in ventilation design is now that it is not the PM mass concentration (which can be measured or at least estimated); it's the visibility which is the decisive factor. While for exhaust PM a correlation between PM mass concentration and visibility exists from experiments made in 1964 (2) and confirmed for diluted exhaust fumes in road tunnels in 1977 (3), no equivalent correlation value existed for non-exhaust PM. In order to tackle this problem one has to deal with a system with multiple unknown parameters which are: fraction of exhaust PM (heavy duty vehicles and passenger cars), fraction of non-exhaust PM (heavy duty vehicles and passenger cars), correlation between exhaust PM and visibility, correlation between non-exhaust PM and visibility. While the first parameter (exhaust emissions for passenger cars (PC) and heavy goods vehicles (HGV)) can be determined with a higher degree of accuracy, knowledge of the second parameter (non-exhaust emissions) is relatively poor. The third parameter (correlation between PM-exhaust mass concentration and visibility) can be taken as well known (assuming that there was no change in the composition of PM exhaust between the 60's and now - which is a very weak assumption). The forth parameter (correlation between nonexhaust PM mass concentration and visibility) was not known. In an attempt to solve the unknowns, an intensive measurement programme was set up in various road tunnels in Austria (longitudinally and transversely ventilated). The data were than analyzed and the respective correlation factors derived. The overall correlation factor for PM in tunnel air has already been implemented in the new PIARC emission factor tables (5). The background and details are presented in this paper. The second part of the paper deals with the problem of visibility in the case of simulations of fire incidents. In many cases numerical simulations are used to demonstrate the smoke movement and the resulting effects of bigger fires. Numerical simulations are based on mathematical equations describing the aerodynamics and the combustion process. However, there is no equation which allows for a description of the visibility within the smoke layers in a tunnel. Based on a relatively simple assumption about the fraction of combustible material in the air a factor was derived which marks the boundary between visible and invisible ranges within the smoke (i.e. in which zones of the fire/smoke a visibility is possible and in which zone no visibility exists). This factor has been derived during fire tests in an Austrian road tunnel comparing the results of the numerical simulation and video detection, and applied and validated in other fire tests which have to take place regularly in each road tunnel before going into operation.