Cloud point extraction (CPE) with inductively coupled plasma mass spectrometry (ICPMS) was applied to the analysis of zinc oxide nanoparticles (ZnO NPs, mean diameter ~40 nm) in water and wastewater samples. Five CPE factors, surfactant (Triton X-114 (TX-114)) concentration, pH, ionic strength, incubation temperature, and incubation time, were investigated and optimized by orthogonal array design (OAD). A three-level OAD, OA(27) (3(13)) matrix was employed in which the effects of the factors and their contributions to the extraction efficiency were quantitatively assessed by the analysis of variance (ANOVA). Based on the analysis, the best extraction efficiency (87.3%) was obtained at 0.25% (w/v) of TX-114, pH = 10, salt content of 15 mM NaCl, incubation temperature of 45 °C, and incubation time of 30 min. The results showed that surfactant concentration, pH, incubation time, and ionic strength exert significant effects on the extraction efficiency. Preconcentration factors of 62 and 220 were obtained with 0.25 and 0.05% (w/v) TX-114, respectively. The relative recoveries of ZnO NPs from different environmental waters were in the range 64-123% at 0.5-100 μg/L spiked levels. The ZnO NPs extracted into the TX-114-rich phase were characterized by transmission electron microscopy (TEM) combined with energy-dispersive X-ray spectroscopy (EDS) and UV-visible spectrometry. Based on the results, no significant changes in size and shape of NPs were observed compared to those in the water before extraction. The extracted ZnO NPs were determined after microwave digestion by ICPMS. A detection limit of 0.05 μg/L was achieved for ZnO NPs. The optimized conditions were successfully applied to the analysis of ZnO NPs in water samples.