With large amount of datapoints I found that importing it as binary files is the fastest, probably mainly because it can directly read it from the files and put it directly in the memory.

For using the dat files you can read it in many different ways:

Going from SemanticImport, to Import, to ReadList you go from slow, normal, to fast. But they have their differences with respect to robustness.

In the end it really depends also on WHAT you read. a 2.4GB file is quite hefty, but the data might be much smaller when stored binary...

Unfortunately, the x and y coordinates seem to be at random rather than on a regular grid.

Sander. Thank you. However, when I tried to save the file, having read it in after a couple of hours, in binary, I got a message to say that it could not force the file to be binary. The file turned out to be 48 million records of 4 numbers, I resampled the data to get the number down below 1million, which it did quickly, and tried plotting that. ListPointPlot3D worked but when I tried ListPlot3D it eventually crashed. I would like to get a 3D view that I could the rotate and look at from different angles. Is there a more efficient way?