We present an approach for electrode modification by using the oxidoreductase cellobiose dehydrogenase from the ascomycete Corynascus thermophilus (CtCDH). CtCDH is a two‐domain enzyme, in which the catalytic dehydrogenase domain (DHCDH) hosts flavin adenine dinucleotide (FAD) as cofactor and is connected through a flexible linker to a small cytochrome domain with a heme b cofactor (CYTCDH). This domain is responsible for the electron transfer from DHCDH to macromolecular electron acceptors, and is capable of direct electron transfer (DET) with electrode surfaces. CtCDH is optimal at pH values between 7 and 9 and exhibits one of the lowest apparent KM values for glucose (2.4×104 μM) in contrast to the majority of other CDHs, which have acidic optimal pH values and have very low or no activity for glucose. Glassy carbon (GC) electrodes were modified by drop‐casting single‐walled carbon nanotubes (SWCTs) and further modifying with aryl diazonium salts (DS). When adsorbed on such GC‐SWCT‐DS electrodes, CYTCDH showed efficient DET in the presence of a substrate. Six different functional groups for the aryl amines were studied and compared to non‐DS modified GC‐SWCT electrodes. The charge of the grafted aryl amines was investigated and it was found that surfaces modified with aniline, 4‐aminophenol, and 4‐aminobenzoic acid (ABA) contain negative charges at pH 7.4. On the other hand, surfaces modified with p‐phenylenediamine (PD), N,N‐dimethyl‐p‐phenylenediamine, and N,N‐diethyl‐p‐phenylenediamine remain uncharged. To date, the highest DET current density (JMAX=32.5 μA cm−2 for lactose and JMAX=16.2 μA cm−2 for glucose) for a CDH‐modified electrode at human physiological pH can be obtained by using a GC‐SWCT‐ABA CtCDH modified electrode. The use of glutaraldehyde (GA) in GC‐SWCT‐PD‐modified electrodes increases the JMAX twofold. The modified electrodes with DS showed a more negative onset potential for the catalytic current. For GC‐SWCT‐ABA modified electrodes and GC‐SWCT‐PD with GA, a change in the onset potential of −62 and −66 mV, respectively, was found compared with non‐DS modified electrodes. The prepared bioanodes show a loss in response current of 15 % after 9 days of continuous measurements from the original signal in 5 mM glucose, using 50 mM phosphate buffer solution at pH 7.40 at 37 °C.