Gender bias against girls in nineteenth-century England has received much interest but establishing its existence has proved difficult. We utilise data on heights of 16,402 children working in northern textile factories in 1837 to examine whether gender bias was evident. Current interpretations argue against any difference. Here our comparisons with modern height standards reveal greater deprivation for girls than for boys. Discrimination is measured in girls' height-for-age score (HAZ) falling eight standard errors below boys' at ages 11, 11.5 and 12 years of age, capturing the very poor performance of factory girls. But this result cannot be taken at face value. We query whether modern standards require adjustment to account for the later timing of puberty in historical populations and develop an alternative. We also test the validity of the age data, considering whether parents were more prone to lie about the ages of their daughters, and question whether the supply of girls was fundamentally different from that of boys. We conclude that neither proposition is justified. Disadvantage to girls remains, although its absence amongst younger children precludes an indictment of culturally founded gender bias. The height data must remain mute on the source of this discrimination but we utilise additional information to examine some hypotheses: occupational sorting, differential susceptibility to disease, poorer nutrition for girls, disproportionate stunting from the effects of nutritional deprivation, and type and amount of work undertaken. Of these we suggest that girls had to do arduous physical labour in the home alongside their factory work. The only (unsubstantiated) alternative is that girls were more likely than boys to be put into factory work below the legal age limit. Both represent forms of gender bias.