For Blocked-to-Blocked and Unblocked-to-Unblocked transitions
Vernier photogates have an electronic timing resolution of 1 microsecond. You will, however, typically get measurements with as much as +/-25 millisecond variation. This is due to variations in optical-response tolerances of the photodiodes and the geometry of the object relative to the path it takes through the photogate.

When using multiple Go Direct Photogates connected wirelessly, you will typically get measurements with +/- 35 millisecond variation. This 10 millisecond difference is associated with the initialization of multiple device clocks when starting data collection. Two Go Direct Photogates daisy-chained together do not experience this added error as they are utilizing a single clock. (Daisy-chaining two Go Direct Photogates requires a Go Direct® Photogate Timing Cable (VPG-CB-GDX), sold separately.)

For Blocked-to-Unblocked and Unblocked-to-Blocked transitions
Due to differences in the optical-response of a photodiode when transitioning from unblocked-to-blocked compared to blocked-to-unblocked, photogate timing modes that utilize both blocked and unblocked events (for example, Gate Timing and Gate and Pulse Timing) will give significantly higher error in the timing measurements. Whenever possible, avoid using these timing options for data collection.

For a good discussion of these issues, see “Photogates: An instrument evaluation,” Eugene P. Mosca and John P. Ertel, Am. J. Phys. 57 (9), 840-844 (1989).