Sunflower and rapeseed meals are agro-industrial coproducts that contain high amount of phenolics (1-4 % dry matter), mostly as esters of caffeic acid (CA) and sinapic acid (SA), respectively. The enzymatic hydrolysis of the ester bonds enables to recover the corresponding free phenolic acids that are bioactive compounds and platform molecules for various applications in green chemistry. Here we aimed to find the best route for producing free CA and SA by applying various fungal carboxylic ester hydrolases from recombinant Aspergillus niger strains either directly on crude meal or on their phenolic extracts obtained by methanolic extraction. Two types of meals were studied: (i) industrial (commercial) meals (I-meals), produced by a process that includes cooking at 95-100°C and steam desolventizing at 105-107°C, and (ii) non-industrial meals (NI-meals) obtained at pilot-scale with much milder heat treatment, that offer a higher total phenolic content. CA release through hydrolysis of sunflower meal (SFM) was successfully achieved with A. niger type-B feruloyl esterase (AnFaeB) and chlorogenic acid esterase (ChlE). Maximal amount of free CA released was of 54.0 ± 1.1 to 59.8 ± 2.1 µmol/g defatted dry matter (DDM) from I-SFM (94-100% hydrolysis yield) against 42.0 ± 1.1 to 52.3 ± 0.2 µmol/g DDM (59-73% hydrolysis yield) from NI-SFM in which CA release was hampered by a phenolic oxidation side-reaction, seemingly due to meal endogenous polyphenol oxidase activities. AnFaeB and ChlE hydrolysis of phenolic extracts from NI-SFM increased the CA amount obtained to 55.0-68.1 µmol/g DDM (77-95% hydrolysis yield). In all cases, AnFaeB showed broader specificity towards SFM caffeoyl quinic acid isomers than ChlE. In particular, ChlE did not hydrolyze 3-O-caffeoylquinic acid. The maximal amount of free SA released by AnFaeA hydrolysis was 41.3 ± 0.3 µmol/g DDM from NI-SFM (50% hydrolysis yield) and 32.3 ± 0.4 µmol/g DDM from the phenolic extract (64% hydrolysis yield), with AnFaeA also having sinapine transesterification activity that led to the synthesis of 1,2-di-Osinapoyl-β-D-glucose. Finally, of all the substrates tested for enzymatic hydrolysis in our conditions, I-RSM and NI-SFM extract showed the best compromise between initial total phenolic content, hydrolysis yields and amounts of CA/SA released.