The use of cement in the building industry accounts for a large part of the world's CO2-emissions. In the developing part of the world where materials are costly and labour is cheap, cement is expensive. In Nepal, a country which is frequently hit by earthquakes, low-cement-content building blocks are already being made and research on how to reduce the cement content further is currently going on. The manufacturing process for these blocks varies significantly, so a method to predict block strength easily and implement cost-effective quality control could substantially enhance production quality. The purpose of this study is to investigate how to predict the strength of building blocks using the water content, cement content and bulk density. Additionally, it examines whether bulk density could serve as a reliable indicator of strength, which could simplify quality control by allowing blocks to be weighed. This, together with an evaluation of the production site performance, is used to calculate the theoretical improvement potential and a proposed cement reduction. Through experiments, the compressive strength was measured for blocks of mixes with different cement percentages under various water content conditions, resulting in blocks with differing bulk densities. This data was used to create a model that predicts the compressive strength. Field visits to five sites were conducted to evaluate the current performance and this data was used to calculate the improvement potential. The results show the importance of water for the workability of the mix and in turn the compressive strength of the blocks. More water and more material into the mold lead to an increase in the strength and the theoretical improvement potential for cement reduction is 50\% by only these measures. The bulk density is a promising indicator of the compaction and compressive strength and weighing the blocks at creation could potentially reduce the variability in the strength of the blocks.