The article discusses selected findings of a narrative review, funded by the Institute of Physics, in response to the continuing decline in the number of girls studying physics post-16 in England; 177 selected sources, of national and international research literature about girls' participation in physics, were reviewed. In the article, we argue that gender and science are mutually constitutive and girls' participation in physics education, historically and currently, needs to be understood in relation to this. Prior achievement and perceptions of the difficulty of physics are determinants of students' decisions about whether to continue to study physics. These influences may be heightened for girls by gendered associations about who is, and is not, competent in mathematics and physics. Interest and enjoyment in physics also influence students' course choices, particularly those of girls, and these decline relative to other sciences through schooling, more so for girls than boys. This decline is not disrupted by school organization. Single-sex organization is associated with high teacher expectations in science and a greater sense of "belonging" for girls, but not for all girls. The contents, contexts and ways of approaching problems and investigations in physics more closely reflect what boys, more than girls, engage with outside school, and those activities associated with what culture defines as masculine rather than feminine attributes. These exert a negative influence on girls' engagement with physics, their sense of self-efficacy in relation to it, and their perception of its personal relevance. Girls, relative to boys, continue not to see a future self engaged in physics and physics-related careers. This can be disrupted by changes in the curriculum and in pedagogy. Context-based courses alter how physics content is organized, and impact positively on overall performance, and on girls' performance relative to that of boys. They also raise fundamental challenges to physics education and its perceived educational purpose. Developments in science education in England, we suggest, do not challenge the gender-science relationship and their impact on girls' participation may be limited as a consequence. The article highlights the paucity of UK-based research generally into girls' participation in physics, and particularly into pedagogy that enhances participation for girls from a variety of different ethnic and socio-economic backgrounds. (Contains 1 figure and 6 notes.)