Owton and Allen-Collinson challenge the conventional assumption that emotional involvement contaminates qualitative research. Friendship is examined neither as methodological innocence nor as a licence for intimacy, but as a relation that can generate forms of access, disclosure and embodied understanding unavailable to detached observation. Their key operation is reflexive calibration: proximity must be made analytically visible, including its affects, asymmetries, expectations and ethical limits. The friend-researcher occupies a shifting position in which knowledge is produced through trust but is never guaranteed by it. This reframes bias as a problem of provenance rather than a defect solved by distance. Emotional reflexivity becomes evidentially productive when the researcher records how the relationship conditions what can be asked, seen, remembered and interpreted. The article links ethnography to feminist and relational methodologies, providing a vocabulary for long-term research situations in which fieldwork and ordinary social life overlap. It is especially important for any archive produced through recurring companionship rather than episodic professional access.