This study was carried out to compare key haematological features of diploid (2n) and triploid (3n) turbot as a first step towards the assessment of the ability of 3n turbot to withstand sub-optimal culture conditions. Morphometric indices of erythrocytes were determined on blood smears by light microscopy. Triploidy significantly (P<0.001) increased all morphometric indices measured in the erythrocytes, including size, surface, and volume, except for the size of minor nuclear axis. The increase in cell size was larger for the major (31.0%) than for the minor (8.3%) axis, thus rendering erythrocytes of 3n turbot more ellipsoidal. The increase in erythrocyte volume (45.9%) was close to the theoretical expected 50% increase as a result of one extra chromosome set. Haematological indices were measured automatically by a haematological Coulter Counter. Triploid turbot had lower numbers of red blood cells (RBC: 1.84 cells pL(-1) in 2n vs. 1.27 cells pL(-1) in 3n; P<0.001) but they were of a larger size (Mean corpuscular volume [MCV]: 145.51 fL in 2n vs. 181.78 fL in 3n; P<0.001). However, the decrease in RBC was not compensated by the increase in MCV, and thus, triploidy decreased the haematocrit (Hct: 26.80% in 2n vs. 23.11% in 3n; P<0.001) and total blood haemoglobin concentration (Hb: 73.74 g l(-1) in 2n vs. 67.54 g l(-1) in 3n; P<0.05). In contrast, mean corpuscular hemoglobin (MCH: 40.27 pg in 2n vs. 53.28 pg in 3n; P<0.001) was higher for 3n turbot as a result of their larger erythrocytes although MCH concentration (MCHC: 0.28 pg fL(-1) in 2n vs. 0.29 pg fL(-1) in 3n did not significantly differ. RBC, Hct and MCV were also determined manually using light microscopy. In general, discrepancies between the two methods were small (overall approximately 7%) but the Coulter Counter tended to overestimate RBC and Hct (and thus to underestimate MCV). Nevertheless, relative differences between ploidies were very similar, thus verifying triploidy-associated changes in hematological features. These changes, as determined in the present study, are essential when evaluating the feasibility of triploid turbot for intensive aquaculture systems in which unfavorable situations may occur.