The application of virtual reality (VR) is increasing steadily. However, cybersickness - a diffuse set of symptoms like discomfort and nausea - remains an accessibility problem for VR. The eye movement hypothesis stresses the role of the optokinetic nystagmus (OKN), a distinct eye movement pattern for gaze stabilizing in moving scenes, in the genesis of cybersickness symptoms. Thus, we postulate a time lag between OKN and cybersickness onset, as well as an explorative approach for using the OKN as a diagnostic criterion. For the study, 70 subjects were exposed to two VR environments in a randomized order. One of the VR environments aimed at achieving a naturalistic VR use case scenario while the other used an optokinetic drum in VR for inducing the OKN. Each participant rated cybersickness Pre/Post-VR on a symptom questionnaire and during VR on a single-item questionnaire every two min. To reduce the error probability in velocity-based eye event coding we applied a 3-dimensional vergence-based algorithm for the pre-processing of the eye-tracking data. The results show that cybersickness was successfully induced in both VR environments but with different main symptom facets. However, there was a negative correlation between the frequency of occurrence of OKN and upcoming reported cybersickness. None of the OKN parameters served as diagnostic predictors for cybersickness. We discuss methodological limitations regarding the applicability of physiological indicators for predicting cybersickness and the advantages of openness.