Car repair painters usually experience long-term exposure to many different solvents. In Italy, the greater part of car painting shops are "handicraft", i.e. 2-5 workers and small premises. Usually workers do not have specific duties, but everyone takes part in all different operations. Moreover, working time is not standardized but varies according to the workload, working methods tend to be traditional, and compliance to individual protection devices is poor. We have hence assessed, on a sample of 8 italian handicraft car painting shops, the exposure levels to solvents, implementing three classic exposure monitoring methods: environmental sampling with charcoal tubes, personal sampling with diffusive charcoal samplers, and urinary determination of unmetabolised solvents. A simple regression analysis was performed to evaluate relationships between the three series of data. The solvents analysed were toluene, ethylbenzene, 1, 2-dichloropropane, n-butylacetate, n-amylacetate, xylene isomers, ethylacetate, and benzene. Benzene was included due to several factors, among others its presence (1% v/v) in the italian unleaded gasoline. Benzene was found in all shops, at levels around or higher than the 8-h timeweighted average limit (8-h TLV-TWA). Other solvents were found in the different shops at various levels, 10 2 -10 1 times the 8-h TLV-TWA. Air concentrations of toluene, n-butylacetate, xylenes, and benzene were positively correlated with urinary levels of the parent compounds, while a negative correlation was found for ethylbenzene. The health implications of workers' exposure level was briefly discussed.