Background The use of child restraint systems (CRS) is a passive safety intervention that has proved to be effective in reducing road traffic injuries (RTI) in children. As part of the Bloomberg’s Global Road Safety Program in Mexico, an intervention addressing non-use of seatbelt and CRS was implemented in the cities of León, Guanajuato and Guadalajara, Jalisco. Cuernavaca, Morelos was selected as comparison city. The purpose of this analysis was to estimate the prevalence of CRS use in children ≤5 years of age and to identify factors associated with its use.

Methods We performed direct observation in a randomly selected sample of preschools. We collected information as children were arriving school about demographics, use of CRS, characteristics of the vehicle, the driver and other car occupants. Public transportation vehicles or school buses were excluded. CRS prevalence was calculated for each city. A logistic regression analysis was fitted in order to evaluate factors associated with CRS use in children.

Results Across the two rounds of observations, the prevalence of CRS use was 12.1% (CI: 10.2, 14.3) in León, 17.2% (CI: 15.4, 19.2) in Guadalajara and 19.4% (CI: 16.8, 22.1) in Cuernavaca. Regression analysis showed that factors associated with a higher prevalence of CRS were driver’s seatbelt use (OR: 1.3; CI: 1, 1.6) and driving a van (OR: 1.8; CI: 1.4, 2.2). The probability of using a CRS decreased if the driver was man (OR: 0.6; CI: 0.5, 0.7), if the children travelled in the front seat (OR: 0.1; CI: 0.08, 0.2), in taxi (OR: 0.1, CI: 0.03, 0.3), and if the number of passengers increased (OR: 0.7; CI: 0.6, 0.9) (p<0.05).

Conclusions CRS use is lower than reported in high-income countries. We were able to document the presence of factors that adversely affect the use of CRS. Some of them had been widely documented, as the number of passengers in the vehicle. This information will support the development of targeted interventions to increase CRS use awareness.