The aim of this investigation was to determine the extent of het-erosis and combining ability estimates and their interactions with loca‌tions for yield and yield components in Egyptian cotton. Six parental va‌rieties namely Giza 85 (P1), Giza 86 (P2). Giza 89 (P3), Giza 76 (P4), Gin 77 (P5) and Giza 87 (Pe) representing a wide range of variability in most of the studied traits, were utilized. The six parental varieties were crossed using a client crosses mating design (Including reciprocals) to obtain 30 F1 hybrids in 1996 growing season. In 1997 growing season, two experiments were carried out at both Sakha, Kafr El-Sheikh Govern. orate (L1) and Karr Saad. Damietta Governorate (L2). Each experiment was arranged in a randomized complete block design with four replica‌tions to evaluate all genotypes, which involved six parental varieties and their possible 30 F1 hybrids including reciprocal. Data were recorded on 10 guarded plants randomly sampled from each plot. The studied traits were: seed cotton yield per plant, lint yield per plant. number of bolls per plant. boll weight, lint percentage, lint Index and seed index. The results of the anatyses of variances indicated highly signifi-cant differences among population means for all studied traits at the two locations and combined data. Also, the genotypes by locations in‌teractions were significant or highly significant for all studied traits ex‌cept lint percentage and lint index indicating the presence of interaction between genotype and locations. The results indicated that no single parent was the best for all studied traits. Generally, the variety Giza -86 (P2) was the highest yielding parent _TIvo largest amount of heterosis versus mid-parents was (7.79 %) for lint yield per plant, while the low‌est amount of heterosis was (0.91 %) for number el bolls per plant. HeterOSIS versus better parent was undesirable. The mean squares in the Ft hybrids for general combining ability (G.C.A.) were highly significant for all studied traits. Tests of significant of the mean squares of (G.C.A.) ware larger in magnitudes than those specific combining ability (S.C.A.) and showed highly significant differ‌ences for all studied traits. The estimates values of genetic parameters showed that the magnitudes of dominance genetic variance WO) was larger than thre‌lore additive genetic variance (o2A) and reciprocal variance (02r) for all