**Background:** Despite global progress in child survival, Sub-Saharan Africa still has the highest child mortality rates, with many children dying without reaching a health facility. In Ghana, under-5 mortality is higher in rural areas. Healthcare-seeking behavior is influenced by perceptions of illness cause and socio-cultural factors. *Asram* is a traditionally-defined childhood illness complex in Ghana, linked to spiritual causes and perceived as 'not for the hospital.' This study aimed to understand local concepts of *Asram* and its influence on healthcare-seeking behavior of mothers/caregivers in rural Ashanti, Ghana.
**Methods:** This qualitative study was conducted from October 2019 to February 2020 in two rural communities (Akutuase and Wioso) in the Asante Akim North district. Four Focus Group Discussions (FGDs) were conducted with 47 mothers/caregivers of children under-5, and 22 Key Informant Interviews (KIIs) were conducted with mothers/caregivers of children who had *Asram*, health workers at district, facility, and community levels, and *Asram* healers. Participants were selected purposively and via snowball sampling. Data were analyzed iteratively using a thematic analysis approach. The study adhered to the 32-item COREQ checklist.
**Key Results:** Nine types of *Asram* were described: *Asram boredwo* (emaciation with large head, linked to pregnant women eating roasted plantain or dressing indecently), *Asram ntoos* (tomato-like blisters, linked to pregnant women walking through spider webs or eating squirrel meat), *Asram borfre* (swollen head, linked to eating pawpaw during pregnancy), *Asram mpompo* (rashes/boils all over body), *Asram mpaemu* (skull fissures), *Asram ayamtuo* (diarrhea with bloated stomach), *Asram esuro* (convulsions in children aged 1-5), *Asram nofo-denden* (engorged breasts in mothers), and *Asram pepe* (fast breathing, often resulting in neonatal mortality). *Asram* was perceived to be acquired spiritually and/or inherited. Mothers/caregivers trusted *Asram* healers and preferred to call on them first. Treatments included 'baptism with a chameleon,' wall gecko preparations, herbal bath therapies, and herbal mixtures. Payment was not obligatory; caregivers could give a gift ('aseda'). This preference for *Asram* healers was found to be the main reason for delays in seeking healthcare. Health workers reported that children often presented to facilities only when nearly dying, after unsuccessful treatment by *Asram* healers. Even Community Health Volunteers (CHVs) expressed strong belief in *Asram* and its treatment by healers. One CHV stated: 'Madam, there is Asram! I believe it! As for Asram, I also hold the view that it is not meant to be treated at the hospital, because, they cannot treat it, even some of the nurses understand this.' A nutrition officer noted that children with malnutrition were often first taken to *Asram* healers, where herbs caused diarrhea and further weight loss, leading to late presentation with Kwashiorkor.
**Clinical Implications:** *Asram* represents a significant barrier to prompt and appropriate healthcare seeking for children under-5 in rural Ghana, contributing to preventable deaths. The strong community trust in *Asram* healers, even among some health workers and volunteers, undermines biomedical care. The study recommends that healthcare workers capture *Asram* and other traditionally-defined illnesses in the design of social and behavior change communication interventions targeting pregnant women and caregivers. It also suggests that the Ministry of Health incorporate guidelines for appropriate health-seeking for children under-5 in the Ghana Child Health Policy, and consider engaging *Asram* healers, who themselves proposed recognition by the Ministry of Health.