[en] Planned control rod drop time registration was passed two times a year per reactor unit. In 1992-1993 some control rods at almost all WWER-1000 units exceed the prescribed 4 second time limit. More than 7000 individual control rod time tests were made from the main correction measures. The following conclusions have been made from the current statistics data: (a) The main role of the vibration factor is proven in the fuel assembly (FA) bowing process. The greatest drop times and maximum of bowing values are concentrated at the vibration zone (2-4 FA rows from the reactor partition). The first FA row seems to be stable due to the interaction with the reactor partition; (b) Bowing relaxation will proceed during several fuel cycles (estimated value is 4-6), and depends on previous FA use history. It seems to be proven that previously bowed FAs effect the new FA, so previously bowed FAs are straightened until the middle of the fuel cycle. At some reactor units small drop time reduction is observed up to half of the fuel cycle from the start time values; (c) Control rod drop medium time (t) has almost linear dependence on operation time (τ) (t=k x τ + b]. Estimated by the method of least squares, values of k and b differ from unit to unit and from cycle to cycle. Values of k and b are in following ranges: b=2.0 - 2.6 seconds, k=5-50x10^-4 seconds per effective operation day; (d) Control rod drop time distribution changes through operation time. The position of maximum starts to shift after 240 effective days, and the form of the distribution start to change at the same time. Before 240 effective days, the distribution essentially does not change. To guarantee that the control rod system reliability is now within prescribed limits, we should continue testing. Additional analysis is needed. Test frequency can be reduced to avoid additional unreasonable transients. (authors)