Inbreeding trends in three populations of rainbow trout under selection for over seven generations were investigated. Inbreeding levels across generations were all below 11.5% as estimated from pedigree information and effective population size, 'Ne'. Coefficients of inbreeding estimated from 'Ne' were generally lower than those from pedigree information. The average rates of increase in inbreeding were 2.0%, 0.53% and 1.38% per generation for populations 1, 2 and 3, corresponding to 'Ne' of 25, 94 and 36, respectively. These values were within the acceptable range for avoiding loss of fitness and suggest there is no cause for concern over the current inbreeding trend for all populations. Three sire-dam models: additive (A), additive plus dominance (A+D), and additive plus dominance plus additive by additive (A+D+AA) and a fixed linear model were used to estimate the effect of inbreeding on growth. Estimates of inbreeding depression from the fixed model were substantially lower than those obtained from the mixed models. The A+D and A+D+AA models produced similar estimates of inbreeding depression that were higher than those from the A model. Results suggest that addition of AA effect in the A+D model does not have a significant effect on the estimation of inbreeding depression. Additive, dominance, and additive by additive genetic variance components for growth were estimated using the tilde-hat approximation to REML method. Estimates of heritability were the same for all populations (h2 = 0.20) in the A model. These estimates decreased slightly (range of 0.193 to 0.196) when dominance and additive by additive effects were included in the model. The percentage of total genetic variance accounted for by dominance effects averaged among all populations was 49.6% and 43.4% when calculated under A+D and A+D+AA models, respectively. Most of the non-additive genetic variance for body weight was accounted for by dominance effects. Inflated values for the dominance variance may be due to confounding with common environmental effect due to full-sibs. Averaged among all populations, additive by additive effects accounted for 12% of the total genetic variation. These results confirmed the presence of non-additive genetic variances for growth in rainbow trout.