Background Sporting events are associated with violence, creating societal and healthcare system burdens. However, these associations have not been objectively evaluated in low- and middle-income countries.

Aim To determine whether football events were associated with changes in homicide rates in the city of Cali, Colombia, during 2005–2008.

Methods Football days were those in which the city's professional teams or the national team played nationally or internationally. Daily counts of homicides were obtained from the city's fatal injury surveillance system. Homicide rates were computed using population estimates to adjudicate person-years (PY) of exposure. In stepwise conditional autoregressive negative binomial regressions adjusted for victims' demographics, known socio-political and cultural factors, weekends, holidays, trends, and seasonality, homicide rates during football days were compared against homicide rates during non-football days.

Results There were 1352 homicides during 279 football days and 4767 homicides during non-football days (Rate, 82 and 68 per 100 000 PY, respectively, p<0.001). There was an increased risk of homicides during the 182 days of home games (IRR=1.11, 95% CI 1.02 to 1.22) compared to non-football days. There was an increased risk of homicides during the day following the football games (IRR=1.09, 95% CI 1.01 to 1.17) compared to other days. The outcome of the games was not statistically associated with increased homicide rates (IRR=1.07, 95% CI 0.97 to 1.18).

Significance In Cali, violence prevention efforts should be increased on days of home games and on the day following the football games. The healthcare system should be prepared these days for increased demands on trauma care to prevent detrimental effects on performance and patient care.