Abstract Argyrozona argyrozona is an important component of the South African commercial linefishery. The purpose of this study was to develop a reliable protocol for estimating age and calculating growth. Methods based on both otoliths and mark-recapture data were investigated. Transverse sagittal sections from the Tsitsikamma National Park (TNP) showed clear opaque and translucent growth increments. Marginal growth zone analysis and mark-recapture of chemically tagged fish (Oxytetracycline) revealed that these were deposited on an annual basis: opaque in summer and translucent in winter. The first translucent check was shown by daily increment counts to mark the completion of the first annulus (i.e. not a juvenile/false check). Within reader (between counts) and between readers average percentage error (APE) was 5.3 and 1.8, respectively, showing that readability of carpenter otoliths is high. Comparison between whole and sectioned otoliths showed that the former significantly underestimated the age of fish older than 10 years ( P<0.01). The maximum age obtained from whole otoliths was 16 years while that obtained from otolith sections was 27 years. Based on realistic L ∞ estimate and the distribution of residuals, the three parameter von Bertalanffy growth function adequately fitted the observed length-at-age data derived from otolith sections. Growth rates did not differ significantly between sexes ( f=2.217, P=0.09). Growth of carpenter in the TNP is described by the following equation: L t=623(1− e −0.08(t−1.96)) A large proportion (68%) of the individual growth rates derived from mark-recapture data were below those predicted by the otolith based von Bertalanffy growth model ( P<0.01). This was attributed to the negative influence of external tags as hydroids, frequently occurring on the tags of recaptured fish, were observed to cause severe lesions and in some cases, extensive fin damage. This brings into question the use of mark-recapture studies to calculate growth of some species.