Genetic parameters and (Co)variance component were estimated for birth weight (BW), weight at 120 day (W120d) weight at 180 day (W180d) and yearling weight (YW) for a flock of 2248 Barki sheep (1303 males and 945 females) progeny of 74 unrelated rams over a period of 23 years from 1979 to 2001. Analyses were carried out by REML, fitting an animal model and ignoring or including maternal genetic or permanent environmental effects. Four animal models were fitted for all traits studied, and the best model was chosen after testing improvement in log- likelihood values. The fixed effects in the model were year of birth, sex, type of birth and age of dam. Number of days between birth date and the date of obtaining measurement of each record was used as a covariate. Direct heritabilities estimates were inflated substantially for BW and W120d traits when maternal effects were ignored and vice versa for W180d and YW. Based on the most appropriate fitted model, direct heritability of BW, W120d, W180d and YW were 0.24±0.05, 0.20±0.03, 0.19±0.03, and 0.18±0.02, respectively. Corresponding maternal heritabilities were 0.10±0.02, 0.07±0.02, 0.07±0.02 and 0.07±0.03 for above traits, respectively. Maternal genetic effects contributed 10 to 14% of the total phenotypic variance for BW, and their effect diminished further with other traits. Estimates of the fraction of variance due to maternal permanent environmental effects were 0.07±0.0.02, 0.09±0.0.03, 0.06±0.02 and 0.07±0.02, for BW, W120d, W180d and YW, respectively. These results indicate that selecting for improved maternal and for direct effects in Barki sheep would generate slow genetic progress in growth traits. Direct and maternal genetic correlations (r_a1a2, r_m1m2) among the lamb weights varied between 0.67 and 1.00 and between 0.79 and 0.98, respectively. The results showed that the maternal influence on lamb weights decreased with advanced in age at measurement. More over, ignoring maternal effects from the model caused overestimation of direct heritability. Maternal effects are significant sources of variation for growth traits and ignoring these effects in the model would cause inaccurate genetic evaluation of lambs.