Abstract The routine use of prophylactic antibiotics in neurosurgery has been shown to significantly reduce surgical site infection rates. The documentation of non-surgical site, nosocomial infections in neurosurgical patients remains limited, despite this being a stimulus for prolific antibiotic usage. The actual quantum of antibiotic use in neurosurgery and its role in infection control remain both undocumented and controversial. The authors address this issue with a cost-effectiveness study using historical controls. Bacteriologically positive body fluid samples were used to quantify infection rates in the year 2006 and compared with those in the year 1997. Itemized drug lists obtained from dedicated neurosurgical intensive care units and wards were used to quantify antibiotic usage and calculate their costs. Results were compared using both historical and internal controls. The monetary conversion factor used was INR 40=US$1. A total of 3114 consecutive elective and emergency neurosurgical procedures were performed during the study period. 329 patients (10.6%) were recorded to have bacteriologically positive body fluid samples, and 100,250 units of antibiotics were consumed costing Rs. 14,378,227.5 ($359,455.7). On an average, an operated patient received 32.2 units of antibiotics valued at Rs. 4,617 ($115.4). The crude infection rates were recorded to have reduced significantly in comparison to 1997, but did not differ between mirror intra-departmental units with significantly different antibiotic usage. Antibiotics accounted for 31% of the per capita cost of consumables for performing a craniotomy in the year 2006. This estimate should be factored into projecting future package costs.