This study examines cross-linguistic influence of L1 on L2 and L2 on L1 and the extent to which age of L2 onset (L2 AoO) is linked to the acquisition of morpho-syntactic properties in both languages of bilingual children who acquire L1-Russian as a heritage language and L2-Hebrew as a majority language. Our investigation of L1-L2 influence focuses on morpho-syntactic features, whose configurations vary in Russian and Hebrew. Definiteness is realized in Hebrew (but not in Russian), aspect is selected in Russian (but not in Hebrew), and [ACC] case is realized in both languages (but the mapping is different across the two languages); finally, the features of [Person], [Number] and [Gender] are mapped onto verbal inflections in both languages. A total of 110 Russian-Hebrew bilingual children aged 5;5–6;5 with varying ages of L2-Hebrew onset (0–60 months), 20 Hebrew-speaking monolinguals and 20 Russian-speaking monolinguals participated. Results demonstrate cross-linguistic influence, showing that it is bi-directional (L1 on L2 and L2 on L1). The patterns of cross-linguistic influence were similar: bilinguals performed similarly to monolinguals on features, with similar configurations in L1 and L2 (i.e., subject-verb agreement) but performed lower for properties realized differently in L1 and L2 (i.e., [DEF] articles in L2-Hebrew; [PERF] aspect and [ACC] case inflections in L1-Russian). The results also showed an effect of L2 AoO on the acquisition of both L1 and L2. Children with earlier AoO to L2-Hebrew (before 24 months) achieve better mastery in L2-Hebrew and performed lower in L1-Russian. Conversely, later AoOs to L2, led to better mastery of L1 and weaknesses in the acquisition of L2. Findings are discussed in light of the Feature Re-assembly Hypothesis.