A forest tree growth-response to atmospheric deposition is expected to arise indirectly through soil chemical changes and would probably be observable only in the long term. We examined this hypothesis by evaluating the relationship between periodic height growth of mature northern red oak (Quercus rubra L.) trees and soil, physiography and atmospheric sulfate deposition along a 170-km west-to-east gradient of decreasing sulfate deposition in north central Pennsylvania, USA. Height increments for three common 20-year periods beginning in 1929, 1949 and 1969 were estimated from exponential-monomolecular growth functions fitted to stem analysis data for each of 45 trees in 13 ecologically analogous stands along the deposition gradient. Canonical analysis was used to identify a statistically manageable subset of the original 48 independent soil, site and tree (age, crown width) variables strongly associated with height growth. Predictive models relating total (60-year) and the three 20-year height increments to the reduced variable set plus estimated average sulfate and nitrate deposition were derived by best subsets multiple regression. An inherent spatial gradient of decreasing height growth from western to eastern sites was apparent in even the earliest (1929-1948) increment. This inferred non-deposition-related spatial growth trend was accounted for in the 1949-1968 growth increment by introduction of the earliest (1929-1948) growth increment as a significant covariate in the regression model. The inherent growth largely disappeared by the 1969-1988 period as a probable consequence of converging growth rates reported to occur in oaks after age 60 years regardless of site quality. The 1969-1988 growth increment was not as strongly correlated with site factors as was growth in preceding periods, nor was early growth or sulfate deposition significantly related to this height increment. Growth effects from sulfate deposition, if any, would most likely occur within the recent (1969-1988) increment coincident with the period of naturally decreasing growth rate, when site differences and possibly environmental factors would have less influence on growth. Our results give no indication that wet sulfate inputs are affecting northern red oak height growth across the atmospheric deposition gradient.