The neuroblastoma cell-based assay (CBA-N2a) is widely used for the detection of marine biotoxins in seafood products, yet a consensus protocol is still lacking. In this study, six key parameters of CBA-N2a were revisited: cell seeding densities, cell layer viability after 26 h growth, MTT incubation time, Ouabain and Veratridine treatment and solvent and matrix effects. A step-by-step protocol was defined identifying five viability controls for the validation of CBA-N2a results. Specific detection of two voltage gated sodium channel activators, pacific ciguatoxin (P-CTX3C) and brevetoxin (PbTx3) and two inhibitors, saxitoxin (STX) and decarbamoylsaxitoxin (dc-STX) was achieved, with EC(50)values of 1.7 +/- 0.35 pg/mL, 5.8 +/- 0.9 ng/mL, 3 +/- 0.5 ng/mL and 15.8 +/- 3 ng/mL, respectively. When applied to the detection of ciguatoxin (CTX)-like toxicity in fish samples, limit of detection (LOD) and limit of quantification (LOQ) values were 0.031 +/- 0.008 and 0.064 +/- 0.016 ng P-CTX3C eq/g of flesh, respectively. Intra and inter-assays comparisons of viability controls, LOD, LOQ and toxicity in fish samples gave coefficients of variation (CVs) ranging from 3% to 29%. This improved test adaptable to either high throughput screening or composite toxicity estimation is a useful starting point for a standardization of the CBA-N2a in the field of marine toxin detection.