Knowledge Base Question Answering (KBQA) aims to answer userquestions from a knowledge base (KB) by identifying the reasoningrelations between topic entity and answer. As a complex branchtask of KBQA, multi-hop KGQA requires reasoning over multi-hop relational chains preserved in KG to arrive at the right answer.Despite the successes made in recent years, the existing works onanswering multi-hop complex question face the following challenges: i) suffering from poor performances due to the neglect of explicit relational chain order and its relational types reflected inuser questions; ii) failing to consider implicit relations between thetopic entity and the answer implied in structured KG because oflimited neighborhood size constraints in subgraph retrieval based algorithms. To address these issues in multi-hop KGQA, we proposea novel model in this paper, namely Relational Chain-based Embed-ded KGQA (Rce-KGQA), which simultaneously utilizes the explicitrelational chain described in natural language questions and the implicit relational chain stored in structured KG. Our extensiveempirical study on two open-domain benchmarks proves that ourmethod significantly outperforms the state-of-the-art counterpartslike GraftNet, PullNet and EmbedKGQA. Comprehensive ablation experiments also verify the effectiveness of our method for multi-hop KGQA tasks. We have made our model's source code availableat Github: https://github.com/albert-jin/Rce-KGQA.