Vbus Drop 

USB 2.0 Specification, Section 7.2.2
USB 3.0 Specification, Section 11.4.2
On-The-Go and Embedded Host Supplement to the USB Revision 2.0 Specification, Section 4.2.1
Battery Charging Specification, Revision 1.2, Section 4.2.1
Vbus up to 5.5V ECN

The voltage drop test is a DC measurement that can be done with a regulare DMM by measuring the voltage over Vbus and GND on a fixture. It's however too often products fail these requirements that cause serious problems in the market. The problems only increase now more devices taking more current for charging their batteries over USB. 

For each downstream port Vbus should be measured without load and with all ports loaded with the worst case load. The voltage should never exceed the 5.5V. How mush the load is for each port and how mush the voltage may drop (Vbus min) depend on the following:  

Downstream port under test
Load (mA) Vbus min (V) (*)
USB 2.0 Bus-Powered hub 100 4.4
USB 2.0 Self-Powered hub 500 4.75
USB 3.0 Self-Powered hub 900 4.45
PC USB 2.0 Host system 500 4.75
PC USB 3.0 Host system 900 4.45
Embedded Host & OTG- A USB 2.0 low power (TPL Imax <100mA) TPL (**) 4.4
Embedded Host & OTG- A USB 2.0 high power (TPL Imax >=100mA) TPL (**) 4.75
Embedded Host & OTG- A USB 3.0  TPL (**) 4.45
BC 1.2 CDP capable hub, host embedded or host port 1500 4.75
Drop table

(*) The Vbus min values of the above table are DC values and not take transient voltage into account.
(**) The maximum load for an Embedded host & OTG- A is depending on the TPL. 

When doing the measurement take the cable resistance / voltage drop into account what can be significant with high currents. For example if you have 0.25Ohm resistance for cable and connectors and a current of 900mA you will have a voltage drop of 0.225V. Therefore the measurment should be done as near to the A-Receptacle as possible and if accessible you can measure at the A-receptacle Vbus/GND soldering pad. Measuring at the A-receptacle is the location the USB specifications define to measure but since this often too difficult to access you probably use a fixture and maybe also a cable in between, know that these will give some additional voltage drop.

Keep the following item in mind for drop for testing:
- When measuring Vbus stress the host or hub with compound device in it's worst case power consumption.
- During testing use the power supply that is used in the market and when changing the power supply re-test Vbus drop.
- Many devices on the market take more current then allowed so it's advisable to take more margin and therefore test with loads.
- Many self powered hubs also operate in bus-powered mode but not claim this via their descriptor for those hubs it's not possible for to pass the Vbus drop test when loading multiple downstream ports accordingly to the above table.

More info: 

Maximum allowed worst case scenario within spec:
USB 2.0 

USB 3.0