Background/context: Exposure to community-level violence in childhood is a strong predictor of developmental and cognitive outcomes. Several systematic reviews, as well as meta-analysis have documented how being exposed to violent crimes in developed and developing countries predicts externalizing and internalizing symptoms (Löfving-Gupta, et al., 2018; Flannery, et al., 2007; Gaylord-Harden, Cunningham, & Zelencik, 2011). However, most of the existing evidence that links local violence exposure and children's outcomes explores the observed relation between victimization and development, and thus, results are mostly correlational. A recent body of literature has explored deeper to identify the causal effect of indirect exposure to violent crimes, using geocoded crime information. By exploiting variation in the timing on which crimes occur, researchers can compare exposed and unexposed people that are similar in other observable and unobservable characteristics. This design has been employed to identify the negative effects of violent crimes occurring around school-aged children's homes on their cognitive outcomes (Sharkey, 2010; Sharkey, et al., 2014; Schwartz, et al., 2021; Monteiro & Rocha, 2017), non-cognitive measures (McCoy, Cybele Raver, & Sharkey, 2015) mental health (Cuartas & Leventhal, 2020; Heissel, et al., 2018), and parental engagement in stimulating activities with their children (Cuartas, McCoy, & Molano, 2018). Sharkey et al. (2012), with a similar design, identifies that preschool-aged children exposed to at least one violent crime present negative effects on attention, impulse control, and language, with larger effects as crimes occur closer to children's home. Consistently, evidence suggests that the negative effects of exposure to acute shocks are detectable if the violent event takes place within 2 weeks prior to the assessment. Purpose/Objective/Research Question: The present study estimates the causal effect of attending an early childhood education center (ECEC) exposed to a violent crime a few days prior to children's assessments in Colombia. Children were assessed twice in 2019 in Emergent Literacy and Language, Emerging Numeracy, Socioemotional Development, Motor Development, and Executive Functions Development. We also explore the effect of these shocks on classrooms' pedagogical quality and activities, as coded from an observation protocol concurrent with children's assessment, as potential mechanisms of how community-level violence affect children. Setting: Colombia is amongst the most violent countries. Colombian cities have been consistently classified among the most violent in Latin America, and by extension of the world (Insight Crime, 2022), as for 2020 Colombia reported the 7 highest homicide rate in the world (The World Bank, 2023). We situate our study in this setting to understand the effects of day-to-day violence on ECE in a country where exposure to violence is common, and children are at higher risk of being indirectly exposed to community-level violence, than children living in less violent settings. Population/Participants/Subjects: Participant information comes from a sample of 298 children (Age in months at time 1: M=45.8; SD=4.89), from 51 classrooms in 43 different ECEC. Information was collected in 2019 on centers serving and located in vulnerable communities in two urban settings from Colombia. ECEC in the study share the characteristics of being targeted toward children living in the same neighborhood. Children were assessed in two collections (M=6.63 months between each other) and the quality of the school was coded in the same dates. Research Design: Children are not randomly assigned to crimes, making it difficult to isolate the effects of violence from other environmental risks. We address this challenge with a quasi-experimental design which isolates the effect of recent and atypical exposure to violent crimes on developmental assessments for children attending early childhood centers, as well as process quality indicators as potential channels through which violent shocks can affect children. We provide causal evidence by exploiting variation in the timing and location of violent crimes occurring near public ECEC targeted for vulnerable families in relation to child assessments dates with fixed effect models. Data Collection and Analysis: Our model accounts for the endogeneity of exposure to violence and children's developmental processes, discussed above, by relying on an identification strategy that compares children whose main difference is whether their early childhood center was exposed to at least one homicide few days prior to their assessment, within a radius of 500 meters (m). By including child fixed effects, as well as controlling for time varying student, classroom, and center-level characteristics, including frequency of exposure to homicides in the months prior to the study and structural and process quality indicators, we estimate the causal effect of atypical acute exposure to homicides by relying on differences in the timing of children's assessments. We define exposure to homicides with intervals that range from 1 to 7 weeks before the assessment. Findings/Results: Our results suggest that there is a detrimental negative effect of exposure to homicides around ECEC on children's emerging literacy and language, and numeracy skills, if the crimes take place the week prior to the assessment. Similarly, we identify negative effects on the socio-emotional score if shocks took place up to 35 days prior to assessments. Consistently, we observe a reduction in the classroom's pedagogical quality for a similar amount of time. We observe an unexpected positive effect on executive functions for a similar length, with an unprecise estimate for the first two weeks of exposure. We find null effects on motor development, and the total number of activities in the classroom. Regarding thematic activities, we observe how after the occurrence of shocks there is an increase in language and art related activities. We provide robustness checks for our model's exploring sensitivity of different radius of exposure, as well as timing. We find that our results are robust to different specifications. Conclusion: Consistent with our hypotheses and previous empirical work, we provide evidence of the causal relationship between indirect exposure to homicides on school readiness. Our results show that children exposed a month before their assessment perform lower than their unexposed counterparts on social- emotional development and pedagogical quality. Effects become null with longer time windows (i.e., 6 and 7 weeks).