Farmed fish can be exposed to persistent organic contaminants--such as polychlorinated biphenyls (PCBs), dibenzodioxins (PCDDs), and dibenzofurans (PCDFs)--via feed, this eventually resulting in accumulation levels of health concern. To study the correlation between feed contamination, chemical accumulation in fish muscle (fillet), and chemical depletion, an all-vegetal base (or blank) feed was prepared and fortified with a commercial PCB mixture (Aroclor 1254) and six PCDD and PCDF congeners (namely, 2,3,7,8-T(4)CDD, 2,3,7,8-T(4)CDF, 1,2,3,7,8-P(5)CDD, 1,2,3,7,8-P(5)CDF, O(8)CDD, and O(8)CDF) to reproduce realistic low, medium, and high contamination levels. After a 1-month exposure, trout (Oncorhynchus mykiss) were fed with the blank feed and sacrificed every 0.5 months over a 3-month period from exposure end; fillet specimens were sampled at each time. In all groups, the average fish weight increased linearly through the observation period. The chemical diminishing patterns observed were due to the combined effect of clearance and growth dilution: for 10 PCB and four PCDD and PCDF congeners, patterns were described with an empirical one-compartment (fish muscle) model. The canonical pseudo-first-order kinetic equation used was also modified into the form C=[C(0)exp(-k(C)t)] (m(W)t+1)(-1) to distinguish between the contributions to depletion from clearance, exp(-k(C)t), and growth dilution, (m(W)t+1)(-1). Most mean clearance half-life (HL(C)) estimates appear to be greater than 4 months, in a number of cases reaching magnitudes well over 10 months or even negative, thus clearly indicating a non-negligible contribution from a second compartment. Based on means and their 95% confidence intervals, the depletion HL(D) estimates of the 14 selected congeners seem to be comprised between 1.2-3.4 and 1.0-5.0 months, respectively: these values, accounting for both clearance and growth dilution, provide an indication of the relevance of a blank feed as a management option to reduce the overall PCB, PCDD, and PCDF content in farmed trout. Due to a lack of bioaccumulation, O(8)CDD and O(8)CDF yielded no results for evaluation, whereas for many PCB congeners results were insufficient for empirical modelling.