Background Between 2004–2007 there was a relative reduction of partner physical violence of 50% (17.5% vs 8.5%), in Medellin, Colombia.

Aims/Objectives/Purpose To determine factors associated with intimate partner physical violence decreasing in Medellin Colombia, (2004–2007).

Methods Two household surveys were conducted in Medellin, Colombia (2004 n=5781 and 2007 n=4653). Straus Conflict Tactics Scale was used to measure physical intimate partner violence. We calculated the mean annual change in intimate physical violence, using percentage of aggressors and CI 95%. Logistic regression was used to estimate the relationship between change in physical violence and associated factors.

Results/Outcomes Factors associated with this change were an increase of collective efficacy (B=3.3; SE=1.0), and reduction in the following: past year unemployment (B=−0.06; SE=0.12); perceived neighborhood disorder (B=−1.4, SE=0.66), distrust (B=−2.0; SE=1.1), violence and interpersonal aggression legitimating (B=−1.5, SE=0.12), frustrated expectancies and tolerance of law breaking reduction.

Significance/Contribution to the Field Changes in the social environment contributed to physical intimate partner violence decreasing in Medellin. Promotion of public policies that decrease social factors such as machismo attitude and distrust, and promote collective efficacy and employment, may contribute to reduction of partner physical violence.