A really good anechoic chamber is very expensive but you do not need it, to obtain really good results in case of crossover design. You have to got a big room with damping on the walls, ceiling and floor. In this case you should use time window for measurement.
You can also try outdoor environment in quiet area. In this case you should use higher levels of power during measurement.
One method of measurement in case of two-way box is shown in picture below.
The problem is where to set the microphone. First place should be where your ear will be but the problem is reflected waves. So you have to set the microphone closer to take advantege of time window to avoid reflected waves. On the other hand, placing the microphone too close you will measure speakers SPL at some angle. So you have to find compromise between both issues.
At the end you should to also remember that you have to keep the difference in distance between the microphone and both sound sources the same as in the listening position. You have to consider a possible error in comparison to the wavelength in the crossover operating range.
In case of three-way box or two-way box with more than one the same kind of speaker you can consider use distance shift parameter using simulators. In this case you can shift SPL depending on distance shift to distance ratio.
For measurement you have to connect speakers to measurement set using wires similar to wires which one you are going to use in the finished loudspeaker. Even 0.2 Ω difference can be a accuracy issue!
Simulators