Sampling practices determine the accuracy of blood culture in diagnosing bloodstream infection. The main acute hospital in this study introduced aerobic-only routine blood cultures aiming to increase the volume and number of aerobic samples. At the smaller acute site, aerobic–anaerobic pairs were sent routinely. Culture yield and sampling practices were compared at these two sites and it was found that anaerobic cultures increased the yield of pathogens including facultative anaerobes. Volume cultured and number of samples sent fell short of national recommendations. The aerobic-only policy did not result in more blood being cultured. Based on these findings, the main acute hospital is reintroducing aerobic–anaerobic pairs for routine culture.