We conduct high-frequency panel surveys to investigate poverty dynamics, encompassing monetary and non-monetary dimensions, using various consumption and nutritional indicators. These surveys are carried out on random samples of rural households in Uganda, Ethiopia, and Bangladesh. Our findings reveal that a significant proportion of households in the lowest quartile in all three countries remain there after 2-3 months and even one year later. Our analysis using multinomial models suggests that natural shocks increase the likelihood of experiencing poverty in Ethiopia and facing food poverty in Uganda. Additionally, conflict- related shocks are strong predictors of chronic and transient monetary poverty in Uganda and escalate the probability of falling into food poverty in Ethiopia. We also observe substantial adverse effects of economic shocks on food poverty in both Uganda and Ethiopia. Furthermore, our results indicate that having a female head of household reduces the likelihood of escaping poverty by up to 14% in Ethiopia and Bangladesh while decreasing the probability of remaining non-poor by 21% and increasing the likelihood of being poor by 12% in Uganda. We recommend targeted interventions, such as investments in human capital, including education, safety nets, and financial policies that empower households to build their asset base, for instance, by acquiring livestock and promoting women empowerment. Such measures are crucial for reducing poverty and enhancing resilience in these communities.