The size and variability of jury damage awards in tort cases has been a contentious issue for over a decade. Nevertheless, there has been little empirical work addressing the relationship between the size of jury awards to products liability victims and their compensable losses. This paper examines this relationship using a sample of California cases. A compensatory model of damages based on plaintiff, injury, and case characteristics explains a large proportion of the variance in damage awards suggesting that juries rely heavily on monetary and non-monetary loss estimates in determining damages. The evidence argues against charges of unpredictability, and indicates undercompensation on average.