We described the proof-of-principle of a nonoptical real-time PCR that uses cyclic voltammetry for indirectly monitoring the amplified DNA product generated in the PCR reaction solution after each PCR cycle. To enable indirect measurement of the amplicon produced throughout PCR, we monitor electrochemically the progressive consumption (i.e., the decrease of concentration) of free electroactive deoxynucleoside triphosphates (dNTPs) used for DNA synthesis. This is accomplished by exploiting the fast catalytic oxidation of native deoxyguanosine triphosphate (dGTP) or its unnatural analogue 7-deaza-dGTP by the one-electron redox catalysts Ru(bpy)(3)(3+) (with bpy = 2,2'-bipyridine) or Os(bpy)(3)(3+) generated at an electrode. To demonstrate the feasibility of the method, a disposable array of eight miniaturized self-contained electrochemical cells (working volume of 50 mu L) has been developed and implemented in a classical programmable thermal cycler and then tested with the PCR amplification of two illustrated examples of real-world biological target DNA sequences (i.e., a relatively long 2300-bp sequence from the bacterial genome of multidrug-resistant Achromobacter xylosoxidans and a shorter 283-bp target from the human cytomegalovirus). Although the method works with both mediator/base couples, the catalytic peak current responses recorded with the Ru(bpy)(3)(3+)/dGTP couple under real-time PCR conditions are significantly affected by a continuous current drift and interference with the background solvent discharge, thus leading to poorly reproducible data. Much more reproducible and reliable results are finally obtained with the Os(bPY)(3)(3+)/7-deaza-dGTP, a result that is attributed to the much lower anodic potential at which the catalytic oxidation of 7-deaza-dGTP by Os(bPY)(3)(3+) is detected. Under these conditions, an exponential decrease of the catalytic signal as a function of the number of PCR cycles is obtained, allowing definition of a cycle threshold value (C-t) that correlates inversely with the initial amount of target DNA. A semilogarithmic plot of C-t with the initial copy number of target DNA gives a standard linear curve similar to that obtained with fluorescent-based real-time PCR. Although the detection limit (10(3) molecules of target DNA in 50 mu L) and sensitivity of the electrochemical method is not as high as conventional optical-based real-time PCR, the methodology described here offers many of the advantages of real-time PCR, such as a high dynamic range (over 8-log(10)) and speed, high amplification efficiency (close to 2), and the elimination of post-PCR processing. The method also has the advantage of being very simple, just requiring the use of low-cost single-use electrodes and the addition of a minute amount of redox catalyst into the PCR mixture. Moreover, compared to the other recently developed electrochemical real-time PCR based on solid-phase amplification, the present approach does not require electrode functionalization by a DNA probe. Finally, on account of the relative insensitivity of electrochemical methods to downscaling, the detection scheme is quite promising for, use in miniaturized devices and in the development of point-of-care diagnosis applications.