Abstract Two flocks of ewes housed indoors year-round in a controlled environment and bred to lamb alternately at 4-month intervals in an 8-month breeding cycle resulted in 11,060 ewe records. These came from 3528 ewes of the Arcott breeds and their 814 sires, during 24 successive lambings over 8 years (1983–1990). Estimates of genetic parameters were assessed for consistency across February, June or October lambing seasons. Paternal half-sib estimates of heritabilities were 0.08–0.27 for reproductive traits, and 0.26–0.33 for ewe weight at breeding. Corresponding estimates for repeatabilities were 0.08–0.27, and 0.19–0.26. Heritability estimates for total lamb weight per ewe lambing were 0.12–0.28 at birth, 0.05–0.13 at 21 days, and 0.03–0.11 at 91 days. Corresponding estimates for repeatabilities were 0.0–0.08, 0.02–0.20 and 0.05–0.17. Estimates of genetic, phenotypic and environmental correlations between the reproductive traits and ewe weight at breeding were negligible. The genetic correlations between prolificacy and total lamb weight per ewe lambing were 0.63–0.88 at birth, 0.23–0.71 at 21 days, and 0.32–0.95 at 91 days. Corresponding estimates for fecundity were 0.77–0.92 at birth, 0.74–0.87 at 21 days, and 0.85–0.92 at 91 days. Likewise, phenotypic (and environmental) correlations between prolificacy and total lamb weight per ewe lambing were 0.72–0.81 (0.77–0.84) at birth, 0.41–0.51 (0.37–0.54) at 21 days, and 0.47–0.55 (0.41–0.53) at 91 days. Corresponding estimates for fecundity were 0.89–0.93 (0.89–0.94) at birth, 0.73–0.80 (0.71–0.83) at 21 days, and 0.74–0.80 (0.70–0.82) at 91 days. Estimates of genetic correlations among lamb weights at birth, 21 and 91 days were 0.74–0.96. Correspondingly, phenotypic (and environmental) correlations were 0.65–0.93 (0.58–0.93). In general, all estimates of heritability, repeatability and correlation (genetic, phenotypic and environmental) for ewe productivity traits for the Canadian, Outaouais and Rideau breeds were consistent among the February, June and October lambing seasons. There was no evidence to suggest any bias due to the influence of variance associated with additive genetic×season interaction in the parameter estimates. These findings demonstrate the validity of genetic parameter estimates pooled across lambing season and apply to the specific breeding protocol under conditions of 8-month breeding management.