Abstract Measured emissions from building materials (Part I, Nielsen et al., 1996), were evaluated for their sensory effects, odour and irritation, as well as for health effects. The procedures adopted are general. First, if established indoor air standards or guidelines are available, they are to be preferred for the evaluations. Second, if they are not available, odour and irritation thresholds are used. The occupational exposure limits may be used for the evaluation of health effects if applying an additional safety factor between 4 and 40. The actual value depends on the critical effect, but a safety factor of 40 is proposed as a first approximation. Other values must be justified. Third, if occupational exposure limits are not available, two different procedures provide a tentative standard or guideline on the basis of published literature, which of necessity must therefore be collected and evaluated. One procedure estimates the standard from an effect in animals and applies a number of safety factors (each often equal to 10), corresponding to a series of worst-case assumptions. The other procedure evaluates the critical effect and uses fewer specific safety factors to predict the human no-observed-effect level (NOEL). The political safety factor is then determined, i.e. how far below the NOEL the standard or guideline level should be set. The last-mentioned procedure gives a logical concordant system, explaining why different standards or guidelines may be set for outdoor, indoor and occupational exposures, and why exposures exceeding a standard or a guideline need not cause health effects.