A case study using dynamic compaction to improve alluvial soils of a collective housing project area in Carsamba, Turkey, is presented. In-situ field pilot tests were employed to determine the optimum number of tamping, the grid spacing, the effective depth of improvement and the degree of densification of compacted soil. Optimal compaction conditions were found to be: grid spacing - 6.0 m and tamping number under a weight of 15-ton mass falling freely from a height of 18 m - 6.0. Based on the measurements of Standard Penetration Test Numbers (SPT-N) values before and after the dynamic compaction, it was found that SPT-N values were increased by more than 100% near ground surface (including the effect of filling and levelling) at some points, and the depth of ground improvement was experimentally determined to be 9.5 m. Dynamic compaction had an unfavourable effect on thin clay layers of about 0.5 m thick found at varying depths in two different locations since SPT-N values for the clay layers were reduced by as much as 50% after dynamic compaction. However, soil layers beneath the clay layers could be improved to some extent.