Nanosized TiO2 (nanoTiO2) is one of the most widely used nanomaterials, with applications ranging from paints, self-cleaning coatings, pharmaceuticals, to food and cosmetics. In spite of this massive use of nanoTiO2, its biological activity and toxicity remains subject of the intense debate. In particular, there is still considerable uncertainty in the current understanding of the relationship between physico-chemical parameters of nanoTiO2, such as crystalline phase, size, aspect ratio, surface properties, surface defects and surface chemistry and its potential toxicological effects. Motivated by this general problem, this thesis provides a multidisciplinary experimental insight into to the toxicity and photo-toxicity of various forms of nanoTiO2. Firstly, since all of the widely accepted models of nanoTiO2 toxicity involve reactive oxygen species (ROS), electron spin resonance (ESR) in combination with spintrapping was used to measure ROS formation efficiency for the two most industrially important polymorphs of TiO2, anatase and rutile. The study was performed in an unprecedentedly broad range of particle sizes: 3.8 nm to 150 nm and 5 nm to 215 nm, for anatase and rutile nanoTiO2, respectively. Moreover, the photocatalytic and toxic properties of custom-made anatase-based TiO2 nanowires (with a diameter of 35 nm and a length of 0.5-1 µm) were characterized for the first time. For pure anatase nanoTiO2, the maximum ROS generation efficacy was found for nanoparticle sizes in the range of 25 – 30 nm. The ROS generation efficacy of the custom-made TiO2 nanowires was ca. 30% lower, being close to that of the commercial anatase nanoTiO2 with primary grain sizes of 5.3 nm. Secondly, this thesis addressed challenging, complex and still poorly understood processes occurring when nanoTiO2 particles are brought into contact with living cells. In particular, the comparative nanotoxicity study towards human melanoma Lu1205 and WM793 cell lines was performed for three selected nanomaterials: the custom-made anatase-based TiO2 nanowires, the commercial anatase nanoTiO2 with a primary particle size of 5.3 nm and similar in vitro ROS formation efficacy, and the industrial photocatalytic standard, P25 Degussa (a formulation consisting of 80% anatase and 20% rutile, with primary grain sizes of 25 nm). This comparative nanotoxicity study was performed using very low concentrations of nanoTiO2 (2 – 2.5 µg/mL), considerably lower than that applied in the majority of previous ex vivo cell culture studies. A wide palette of spectroscopic and microscopic methods was applied to verify the cellular response to the presence of selected nanoTiO2-based particles, both in the dark and under illumination with UV-A light of low intensity (λ = 365 nm, 1 mW/cm2). In particular, to detect intracellular ROS in cells incubated with the selected nanoTiO2, ESR spin-trapping with an intracellular spin-trap (ACP) and histochemical detection with a ROS-sensitive dye (NBT) were used. These studies were completed by optical microscopy visualization of ROS-induced changes in cell morphology and actin filament organization (FITC-phalloidin staining). Moreover, atomic force microscopy (AFM) was used to follow the changes in the cell topology and elastic properties (AFM force spectroscopy) of cells incubated with nanoTiO2 and exposed to photo-oxidative stress under UV-A illumination. All of these techniques revealed, either directly or indirectly, deleterious effects induced by incubation of living cells with nanoTiO2 and by photo-oxidative stress under illumination with UV-A. Primarily, optical microscopy visualization of selectively stained actin stress fibers and AFM topography images pointed to a rapid disorganization of the actin cortex for cells exposed to photo-oxidative stress in the presence of nanoTiO2. These results corroborated the changes in the cell shape and elasticity measured by AFM. The cell elasticity measurements, quantitatively described in terms of local Young's modulus values, pointed to an internal remodeling of the cytoskeleton in cells exposed to the presence of nanoTiO2 and photo-oxidative stress. The custom made anatase-based TiO2 nanowires revealed generally stronger cytotoxic effects than the anatase nanoTiO2 with a particle size of 5.3 nm, despite having similar photocatalytic properties in vitro. These observations were confirmed by AFM measurements of the cellular elasticity, histochemical detection with a ROS-sensitive dye (NBT) and by intracellular detection of ROS with ESR spin-trapping. The latter technique provided also a new important observation pointing to the enhanced ROS formation in cells incubated with nanoTiO2 in the dark. Overall, the findings stemming from cell toxicity studies were much more complex than those obtained from measurements of ROS generation efficacy in the presence of nanoTiO2 in vitro. In contrast to a marked evolution of ROS generation efficacy as a function of the primary grain size in vitro, the cellular work pointed to the lack of direct and simple correlations between the physical parameters of nanoTiO2, such as particles size, their aspect ratio, and aggregation state, and cell toxicity.