Many languages are spoken across the world which can bring communication gaps where twornpeople that speak different languages cannot communicate. Usually, this communication gap isrnsolved by using a human interpreter. However, the use of human interpreters is expensive andrninconvenient. Many researches are being done to resolve this problem using machine translationrntechniques. Machine translation is an automatic translation of a source language to a targetrnlanguage. This can be speech to speech or text to text translation. In this work, a bi-directionalrntext based machine translation for English and Afan Oromo languages pair using convolutionalrnneural networks is proposed. We started our study with objective of improving the previous workrnon English to Afan Oromo machine translation by making the translation bi-directional byrnapplying convolutional neural network on translations between these language pair. In order tornachieve our objective, we collected parallel corpus data from different sources and divided intorntraining and testing sets. We have used 80% of total dataset for training and 20% of total datasetrnfor testing. Three systems were implemented where the first system uses a word based statisticalrnapproach that used as a baseline, while the second system with recurrent neural networkrnapproach is used as a competitive model and lastly, the third system with convolutional neuralrnnetworks for the bi-directional translation between Afan Oromo and English languages. rn After training and testing these systems on corresponding training and testing datasets, thernconvolutional neural network achieved 3.86 BLEU score improvement on translation fromrnEnglish to Afan Oromo and 3.32 BLEU score on translation from Afan Oromo to Englishrntranslation than baseline system. Also convolutional neural network approach has shown anrnimprovement of 1.58 BLEU score on translation from English to Afan Oromo and 1.51 BLEUrnscore on translation from Afan Oromo to English translation than recurrent neural networkrnapproach. The convolutional neural network approach is faster on training than recurrent neuralrnnetwork approach.