Testate amoebae are valuable indicators of peatland hydrology and have been used in many palaeoclimatic studies in peatlands. Because the species' ecological optima may vary around the globe, the development of transfer function models is an essential prerequisite for regional palaeoclimatic studies using testate amoebae. We investigated testate amoebae ecology in nine peatlands covering a 250-km north-south transect in south-central Alaska. Redundancy analysis and Mantel tests were used to establish the relationship between the measured environmental variables (water-table depth and pH) and testate amoebae communities. Transfer-function models were developed using weighted averaging, weighted average partial least squares and maximum likelihood techniques. Model prediction error was initially 15.8 cm for water-table depth and 0.3 for pH but this was reduced to 9.7 cm and 0.2 by selective data exclusion. The relatively poor model performance compared with previous studies may be explained by the limitations of one-off water-table measurements, the very large environmental gradients covered and by recent climatic change in the study area. The environmental preferences of testate amoebae species agree well with previous studies in other regions. This study supports the use of testate amoebae in palaeoclimate studies and provides the first testate amoebae transfer function from Alaska.