Frequent service delivery interruptions at the public hospitals due to inadequate resources has led to spread of communicable diseases, disabilities and death among patients seeking medical attention. The objective of the study was to determine the relationship between information systems infrastructure available at the hospital warehouse and unit stores for capturing health commodities consumption data and service delivery at level 4 public hospitals in western Kenya. The study was anchored on two theories; Contingency theory and Theory of Constraints. The study adopted cross sectional research design with a target population of 99 level public hospitals and 426 respondents from service delivery units plus management. Quantitative data was collected using questionnaires and observation checklists while qualitative data was collected using interview schedules. Both reliability and validity of the study tools were tested to ensure internal consistency of the tools. Stratified sampling technique was used for selecting the hospitals where sample size was calculated using Yamane formula giving 79 hospitals. Purposive sampling techniques was used to give a sample size of 316 respondents. Data was analyzed using descriptive statistics and inferential analysis. Correlation test was used to test the association between the study variables. The results established significant positive relationships between information systems infrastructure and service delivery where the correlation coefficient indicated that 86.7% variation in service delivery was determined by information systems infrastructure at p=0.000. Also regression coefficient (β) was at 1.134 indicating that a unit change in information system infrastructure led to an increased improvement of service delivery by 1.134 units at the level 4 public hospitals. The study demonstrated that appropriate use of the available information system infrastructure to collect health commodities consumption data promotes health commodities availability and accessibility at the service delivery units thus efficient service delivery to patients. The study recommended need to train health commodities users on operations and appropriate use of the available information systems infrastructure to capture accurate consumption data at the hospitals and also interlinkage of available automated systems to various service delivery units for efficient operations.