The case study company for this thesis is a company that produces electrical and electronic spare parts for motorcycles. This company has been facing defect problem, whose rate is higher than expected and it keeps increasing at an average rate of one percent a year. As defects lead to a rise in production cost, from reworks and scrap, Lean Six Sigma (LSS) approach has been adopted as a research tool, where we focus only on two main products of the company, that is, Models A and B. In this setting, the applied LSS approach is typically a combination between DMAIC framework of Six Sigma and lean concepts, whose aim is to reduce defects from the production processes. In the first phase, i.e. Define phase, the problem and goal are set for each product model – the defects must be reduced below one percent and its sigma level must be at least at the standard level. Then, in Measure phase, all related information is collected and presented in the form of the Modified Value Stream Mapping (MVSM). Based on the constructed MVSM, the problems have become more visible, with all revealed pre-improvement metrics. In this pre-improvement stage, the defective rates are identified at 2.07%, or equivalently 3.54σ, for Model A and 3.5%, or equivalently 3.31σ, for Model B; whereas, the standard defective rate from the companies within the same industry is 4σ. Once all decisive information is revealed, root causes are identified by means of lean tools and techniques in Measure phase. Based on our analysis, there are four main causes and nine sub-root causes in total. However, with Failure Mode and Effect Analysis (FMEA), only five sub root causes are diagnosed as critical. This leads to three main areas of improvement; (i) the development of system and working culture to prevent malpractices, (ii) the improvement of inspection processes, and (iii) the adjustment of instructions for manufacturing processes. Lastly, in Control phase, we introduce an np-chart to continually monitor the production from batch to batch. We also suggest the company to build working culture enhancing strong relationships among organisational members. The results from the improvement for the selected two models show significant defect reduction, and so increase the production’s sigma level. For Model A, the defective rate drops to 0.57%, or equivalently 4.03σ. This defective level surpasses both company goal and standard sigma level of the industry. For Model B, however, the defective rate falls short of its sigma level’s goal of 4σ, where we achieve the sigma level at 3.87σ, or 0.9%. This could be explained by Model B’s complexities, which require skilled assemblers. Additionally, as we introduce new inspection procedures and adjustment, it would take some time before the assemblers get used to the new instructions.