A whole-cell DHA-rich algal product (A-DHA, provided by Evonik Industries) that is rich in DHA (125 mg DHA/g dry matter) is a possible replacement for fish oil in salmonid diets. The nutrient digestibilities of the algal product were measured in rainbow trout in freshwater and in Atlantic salmon in saltwater (32-33 ppm). In experiment 1, rainbow trout (initial weight ~ 300g) were randomly assigned to 12 x 120 L tanks (n = 10 per tank). A reference diet containing 1% Celite as an indigestible marker and three test diets with increasing percentage of A-DHA substitution (6.67%, 13.33% and 20%) were fed. Feces were collected using a settling column and feed and feces analyzed for digestible dry matter (DM), gross energy (GE), ash, crude protein (CP), essential amino acids and total lipid. The digestibility of six long-chain fatty acids including 18:1n-9 (OA), 18:2n-6 (LA), 18:3n-3 (ALA), 20:4n-6 (ARA), 20:5n-3 (EPA) and 22:6n-3 (DHA) was measured. In experiment 2, Atlantic salmon (~170g) were randomly distributed to 12 fiberglass tanks (600L) with 106 fish per tank. The fish were assigned to four diets with the same levels of A-DHA inclusion as for rainbow trout and yttrium oxide (Y2O3) was used as an inert marker. Feces were collected by stripping and the digestibilities of DM, CP and lipid as well as OA, LA, ALA, ARA, EPA and DHA were determined. In experiment 1, the apparent digestibility of dietary DM, GE and lipid in rainbow trout declined significantly with increasing inclusion of A-DHA (P < 0.01). The inclusion of A-DHA had no effect on the digestibility of CP and ash as well as the availability of essential amino acids (P > 0.05). Furthermore, increased inclusion of A-DHA resulted in significantly lower digestibility of ARA, EPA and DHA (P < 0.05). A similar pattern was seen in the digestibility of OA, LA and ALA, although the effect of A-DHA inclusion was not statistically significant. Regression analysis revealed that nutrient contribution from A-DHA had significantly negative linear and quadratic effects on the apparent digestibility of DM, GE, and lipid. The inclusion levels of A-DHA had both significantly negative linear and quadratic effects on digestibility of LA and ALA, whereas only significantly negative linear effect was found on OA. Significantly negative linear and quadratic regressions were observed for the digestibility of ARA, EPA and DHA. The linear regression for CP was significantly negative and the regressions for the individual amino acids were not significant (P > 0.05). In experiment 2, dietary inclusion of A-DHA had a significantly negative effect on lipid digestibility in Atlantic salmon, at all inclusion rates whereas the significant negative effect on digestibilities of DM and CP was only observed in fish fed 20% A-DHA. The digestibilities of OA, LA, ALA and EPA were greater than 91%. In contrast, the apparent digestibilities of ARA and DHA decreased significantly with increasing substitution of A-DHA (P < 0.01). Significantly negative linear and quadratic regressions were found between nutrient contribution from A-DHA to the diets and apparent digestibility of DM, CP and lipid, so were LA, EPA and DHA. However, there were only significant quadratic regressions for OA, ALA and ARA, but not significant linear effects. Subsequently, a twelve-week feeding trial in rainbow trout was conducted to investigate the impact of replacing fish oil with A-DHA in canola-oil-based diets on the growth performance and fatty acid composition and retention. Four experimental diets containing only canola oil (CO; 13.5%), fish oil (FO; 13.5%), canola oil and fish oil (C+F; 7.4% and 6.1%, respectively) or canola oil and A-DHA (C+A; 15.5% and 6%, respectively) were formulated to contain 386.2 g/kg digestible crude protein and 17.58 MJ/kg digestible energy. In addition, the C+A diet was formulated to have the same DHA concentration as in the C+F diet. Each diet was fed to three tanks of rainbow trout (average initial weight of 70g; n = 17/tank) and the fish were fed to apparent satiation 2 times daily. At the end of the growth trial, all fish approximately tripled their weight. No significant differences were noted between the dietary treatments in growth performance as measured by final weight, average weight gain, feed intake, specific growth rate (SGR) and feed conversion ratio (FCR). Although FO and C+A fed fish tended to accumulate more lipids, final whole body lipid content did not differ significantly between dietary treatments (P = 0.11). The concentrations of EPA, DHA as well as total n-3 fatty acid were significantly higher in fish fed the FO diet than fish fed the other 3 diets. The C+A fed fish had lower EPA and higher DHA concentrations compared with the CO and C+F fed fish; however, the differences were not significant. Apparent retention of total lipid in the trout was not significantly influenced by treatments (P > 0.05). Similarly, dietary treatments had no significant effect on the apparent retention of total saturated fatty acids, total mono-unsaturated fatty acids, n-3 polyunsaturated fatty acids and n-6 polyunsaturated fatty acids. The retention of 18:4n-3 (SDA) was significantly higher (> 100%) in fish fed CO and C+A compared with fish fed FO and C+F (< 51%), indicating greater bioconversion of ALA to SDA in the CO and C+A fed fish than in FO and C+F fed fish. The retention of EPA in the CO and C+A fed fish was over 100%, suggesting a net synthesis of EPA in these treatment groups. In contrast, the EPA retention in the FO and C+F fed fish was 55 and 21%, respectively, which showed a tendency to be significantly lower than that in the other two groups (P = 0.09). The CO fed fish had significantly higher DHA retention than fish fed the other 3 diets. The DHA retention in the FO fed fish (112%) was numerically but not significantly higher than in the C+F (66%) and C+A fed fish (73%). Thus, feeding the C+A to rainbow trout resulted in DHA retention equal to feeding the C+F.