The denomination of "accessory peroneal muscle" (APM) refers usually to two muscles: the peroneus quartus (PQua) and the peroneus digit quinti (PQui) which were believed to be variants of a same muscle. Their morphology and prevalence show high variation in relevant literature mainly owing to the diverse terminology used to describe this muscle group. The aim of this meta-analysis is to generate more accurate description and frequency of those muscles. A total of 46 studies (3,928 legs/ankles) met the inclusion criteria. The aggregate results were: (a) a true APM prevalence of 16 % with 10.2 % for PQua vs. 34.3 % for PQui; (b) a crude prevalence of 16.6 % for PQua vs. 21.5 % for PQui; (c) a bilateral prevalence of 1.73 % for PQua vs. 12.5 % for PQui; (d) the PQua was significantly more prevalent in Indian populations compared to all other ancestries, and the PQui is significantly more prevalent in Europeans and Americans compared to Japanese and Korean populations; (e) though a tendency for higher frequency was found in males and on the right specimens, no significance was found for gender and side; for (f) the "surgical" occurrence of PQua in studies dealing with peroneal tendon surgery was 5.5 % where peroneal tendon pathology seems to be not associated with the presence of an APM; (g) the MRI prevalence of APM was 10.6 %; (h) APMs took origin from peroneus brevis in 60 %, from the distal fibula in 36 % and from other structures in 4 %; (i) APMs took insertion on retrotroclear eminence of the calcaneum in 53.4 %, on peroneal trochlea in 6.6 %, merged with a peroneal tendon in 9.3 %, and inserted on the extensor apparatus of the 5th toe in 18.4 %. Despite the acknowledged limitation owing to the varied terminology used to describe this accessory muscle group, the significant differences found between the aggregate frequency estimates of each muscle do not support the hypothesis that both muscles are variants of a same structure.