The effect of the anti-marine-borer treatment of wood using CCA (a pressure impregnated solution of copper, chromium and arsenic compounds) on non-target fouling animals was investigated. Panels treated to target retentions of 12, 24 and 48 kg CCA m(-3) of wood, together with untreated controls were exposed for 6, 12 and 18 months at coastal sites in Greece, Portugal, France and Sweden. General linear model (GLM) analysis revealed significant increases in numbers of certain fouling organisms (the serpulids Ficopomatus enig-maticus, Hydroides spp., Pomatoceros lamarkii and an unidentified species, three species of spirorbid, and the balanids Balanus perforatus and Elminius modestus) with increase in retention of CCA. The effect of CCA on the numbers of recruits may be due to effects on their settlement and survival, but may also be due to suppression of competitors. Significant differences in settlement density of barnacle spat occurred on newly exposed wood and on wood that had been exposed for 6 and 18 months. The relationships between settlement density and retention could be described by logarithmic curves of the form settlement density = a 1n(l + retention)+b. The effects of CCA on settlement are ascribed either to modification of wood surface chemistry leading to changes in surface charge, the availability of Cu, Cr or As at the wood surface, or to modifications to the microbial film. Barnacle settlement was between 6.5 and 14 times more intense on latewood than on earlywood, an effect that was evident in both untreated and preservative-treated wood.