In the present study, we assessed the psychometric proprieties of the Positive and Negative Affect Scale–Child Version (PANAS-C) in a large sample of Palestinian children (N = 1,376) of different age ranges living in refugee camps. In particular, we used standard confirmatory factor analysis to test competing factor structures for the PANAS-C, with a view to developing a stable version of the instrument, suitable for speedy administration in applied and research settings in the contexts of military violence. Four alternative models of the PANAS-C were evaluated: unidimensional; two-dimensional with independent PA and NA scales and covariance of item-level errors unallowed; two-dimensional with dependent PA and NA scales and covariance of item-level errors unallowed; and two-dimensional with dependent PA and NA scales and covariance of item-level errors. The results of the statistical analysis supported a 20-item measurement model comprising the PANAS-C20 Arabic version for children. The items in this best fitting model loaded on two different and negatively correlated factors. These findings encourage full adoption of the PANAS-C20 as a tool for assessing both PA and NA in Palestinian children living in contexts of warfare.